Abstract— Major depression disorder screening and treatment is a challenge currently for the people around the world followed by appropriate treatment after diagnosis. With a growing number of population, its becoming imperative to use this technology of smartphone in the benefit of patients. Machine learning techniques through mobile apps can be a beneficial diagnostic tool. This review involved checking depression apps available on mobile apps specifically seeking questionnaire-based ones. The review also included analysis of data on depression followed by classification and performing a predictive analysis on it.

The results of this analysis indicate that the depression screening mobile apps which used machine learning technique are based on Cognitive behaviour theory (CBT) and PHQ-9 questionnaire targeting mental health of the patients. They have been designed by clinicians, neuroscientist, computer app developers etc. The review also gives an understanding around the depressed population in terms of the age, sex, ethnicity that must be targeted for future development of the apps and the local languages that it must be available for early diagnoses and treatment. Mobile Apps is proving to play a significant role and carries a huge potential have in disease screening, self-management, monitoring, health education, particularly among younger adults.

Keywords — Depression disorder, Machine learning, Mobile app, CBT, PHQ-9

I.     INTRODUCTION

Depression is a mood disorder that causes a persistent feeling of sadness and loss of interest for a sustained period of time. It affects the way we feel, think and behave and leads to a variety of emotional and physical problems. Its treatable with medication, psychotherapy or both and may be long-term. (Mayo clinic,). Common mental disorders can be either depressive disorders or anxiety disorders. These disorders are highly prevalent in the population and impact the mood or feelings of affected persons and the symptoms can range in terms of their severity and duration. About 300 million people are estimated to suffer from depression globally which is equivalent to 4.4% of the world’s population and this number is

increasing particularly in lower-income countries as the life expectancy is also increasing. This disorder affect people of all ages, from all walks of life, the risk of becoming depressed is increased by poverty, unemployment, life events such as the

death of a loved one or a relationship break-up, physical illness and problems caused by alcohol and drug use.(WHO http://apps.who.int/iris/bitstream/handle/10665/254610/WHO-MSD-MER-2017.2-eng.pdf) WHO/MSD/MER/2017.2

Researchers around the world are investigating more cost-effective ways for screening of ASD, and one of the recent developments is through the use of machine learning, as one of its major applications is fast prediction of outcomes based on past data analysed (Thabtah, 2017).This holds true for Major depressive disorder as well. USA alone has  16 million Americans patients with mental disorder. As per WHO the annual global economic impact on depression is estimated to be $1 trillion.

Depression can occur at any stage of life with similar symptoms, but with different causes and treatment choices. It could be postnatal depression, elderly depression, physical illness, children and adolescents etc.

Supportive counselling can treat milder forms of depression. But specific therapies such as cognitive behaviour therapy (CBT) can be effective for  significant depression.

Machine learning is termed by the type of learning performed by computers and which can produce potentially beneficial information for users as a result of “learning” from large volume of data (Witten, Frank, Hall, & Palestro, 2016). Majority of AI  uses to manage depression falls into three major categories of Virtual Counselling where companies are developing machine learning software to identify depression episodes and provide natural language processing support, patient Monitoring where machine learning would monitor patients, predict and prevent the onset of a mental health crisis, precision therapy where machine learning analytics is used to track, correlate cognitive function, clinical symptoms and brain activity.

Effective treatments for depression are available but less than 50% of those affected in the world receive treatment and this could be attributed to lack of resources, lack of trained health-care providers, and social stigma associated with mental disorders. Inaccurate assessment or misdiagnosis has resulted in antidepressant prescription.

There are effective treatments for moderate and severe depression. Health-care providers may offer psychological treatments (such as behavioural activation, cognitive behavioural therapy [CBT], and interpersonal psychotherapy [IPT]) or antidepressant medication (such as selective serotonin reuptake inhibitors [SSRIs] and tricyclic antidepressants [TCAs]). Psychosocial treatments are effective for mild depression whereas antidepressants is an effective form of treatment for moderate-severe depression. Antidepressant is neither used for treating depression in children and nor the first line of treatment in adolescents and should be used with extra caution.(Mar 2018 WHO facts)

Therefore depression is a complex disorder, challenging to diagnose but can be effectively treated in a timely manner. Machine learning methods utilizing anatomical and physiological data acquired from neuroimaging to create models that can identify depressed patients vs. non-depressed patients are also used to predict treatment outcomes. https://doi.org/10.1016/j.nicl.2015.11.003)

https://www.healthcareitnews.com/news/amazon-web-services-boosts-machine-learning-treat-depression

Can Machine learning tool in depression screening through the use of mobile apps predict depression accurately and if so help patients to seek treatment or provide an ongoing mental health support?

II.   Literature Review

Mental disorders are classified into two main diagnostic categories: depressive disorders and anxiety disorders. These disorders are prevalent and have an impact on the mood or feelings of affected persons. The symptoms can range from mild to severe and duration from months to years. These disorders can be diagnosed easily. Over 300 million people globally are assessed to suffer from depression which is equivalent to 4.4% of the world’s population (WHO,2017). The  number of persons with this common mental disorders globally are  increasing particularly in lower-income countries, because the population is growing and more people are living to the age when depression and anxiety most commonly occurs. (WHO,2017).  Another observation is that depression is more common among females (5.1%) than males (3.6%). Prevalence varies by WHO regions, from as low as 2.6% among males in the Western Pacific Region to 5.9% among females in the African Region. Country-specific estimates can be accessed at http://ghdx.healthdata.org/gbdresults-tool. Prevalence rates vary by age, peaking in older adulthood (above 7.5% among females aged 55-74 years, and above 5.5% among males). This has been observed in children and adolescents below the age of 15 years, but at a lower level than older age groups. Fifty percent of depressed population lives in the South-East Asia Region and Western Pacific Region, reflecting the relatively larger populations of those two Regions (which include India and China, for example). The total estimated number of people living with depression increased by 18.4% between 2005 and 2015; this reflects the overall growth of the global population, as well as a proportionate increase in the age groups at which depression is more prevalent. (WHO,)

 

Aldarwish et al (2017)  The use of Social Network Sites (SNS) is increasing nowadays especially by the younger generations. The availability of SNS allows users to express their interests, feelings and share daily routine. Many researchers prove that using user-generated content (UGC) in a correct way may help determine people’s mental health levels. Mining the UGC could help to predict the mental health levels and depression. Depression interferes the patients ability to work, study, eat, sleep and have fun. Therefore from this SNS data, information can be collected related to person’s mood, and negativism and then classifying them  according to mental health levels. The researchers here have created a model that classify the UGC using two different classifiers: Support Vector Machine (SVM), and Naïve Bayes.

Hilbert et al (2017) have researched on Generalized anxiety disorder (GAD) and Major depression (MD) are very difficult to identify and separate from each other in clinical settings, especially GAD using clinical questionnaire data alone. Study by (  ) suggested that neurobiological biomarkers are useful targets especially cortisol and GM volume data to support diagnostic decisions amongst GAD and MD with the help of machine learning.

Nguyen et al report in Mental health America, (2017) stated that Screening is fundamental to getting treatment as primary care physicians providing usual care miss 30% to 50% of depressed patients and likely fail to recognize many common mental health disorders (Nguyen etal, 2017).This treatment was enhanced when results from a positive screening are graphically represented. PHQ-9 is the most accepted screening taken by users online for depression screen (the Patient Health Questionnaire-9 or PHQ-9). Depression Results by Demographics has indicated alarmingly that youth are at a greater risk out of which female youth (62%) scored higher  than male  youth(52%). It takes 10 years on average between the onset of depression symptoms and when individuals receive treatment. The highest risk population are amongst LGBTQ, students, low income group, Caregivers, new or expecting moms, and veterans etc.

Online resources and tools allows screening  and face to face communication enables  treatment that can complement for mental health concerns. Preventive services and recovery services (e.g. peer services, supportive employment, and supportive housing), would provide support and opportunity for individuals with mental health conditions.

Patel et al ,2015 stated that depression is a complex and clinicians face challenges in accurate diagnosis and effective timely treatment. The development of multiple machine learning methods has help to improve management of this disease. Anatomical and physiological data acquired from neuroimaging created models that can distinguish between depressed patients vs. non-depressed patients and predict treatment outcomes. (Patel etal 2015) have researched on background on depression, imaging, and machine learning methodologies, reviewed methodologies of past studies that have used imaging and machine learning to study depression and suggested directions for future depression-related studies.

Ross et al (2015) stated that an important challenge is in paediatric neuroimaging especially early identification of depression patients at risk. He investigated whether machine learning can be used to predict the onset of depression at the individual level. Ross etal,2015 studied thirty-three never-disordered adolescents (10–15 years old) by structural MRI and monitored the participants for 5 years for emergence of clinically significant depressive symptoms. Support vector machines (SVMs) were used to test whether baseline cortical thickness could reliably distinguish depressed and non-depressed adolescents. Classifier performance accuracies were assessed from subsampled cross-validated classification. Baseline cortical thickness has correctly predicted the future onset of depression with an overall accuracy of 70% (69% sensitivity, 70% specificity; p = 0.021). Findings indicated that cortical gray matter structure thickening predicted the subsequent onset of depression.

Kessler etal in 2016 stated that heterogeneity of major depressive disorder (MDD) illness complicates clinical decision making. Using symptom biomarkers to develop clinical relevance has attained partial success but machine learning (ML) models which were developed from self-reports about incident episode characteristics and comorbidities among respondents with lifetime MDD in the World Health Organization World Mental Health (WMH) Surveys predicted MDD persistence, chronicity, and severity with good accuracy. (Kessler et al,2016). Kessler further stated that model prediction, accuracy was also compared to that of conventional logistic regression models, Area under the curve (AUC) based on ML (.63 for high chronicity and .71–.76 for the other prospective outcomes) was consistently higher than for the logistic models (.62–.70) despite the latter models including more predictors. Results confirmed that clinically useful MDD risk stratification models generated from baseline patient self-reports and that ML methods improved on conventional methods in developing such models.

Dipnall etal (2016) used hybrid methodology for large scale data mining and variable selection to account for missing data and complex survey of key biomarkers associated with depression from a large epidemiological study. Then narrowing from machine learning assisted 21 biomarkers associated with depression followed by traditional logistic regression method a final set of three biomarkers were chosen. The final three biomarkers from the novel hybrid variable selection methodology were red cell distribution width (OR 1.15; 95% CI 1.01, 1.30), serum glucose (OR 1.01; 95% CI 1.00, 1.01) and total bilirubin (OR 0.12; 95% CI 0.05, 0.28). Systematic use of a hybrid methodology for selection of variables selection and  fusing data mining techniques using a machine learning algorithm with traditional statistical modelling, accounted for missing data and complex survey sampling methodology. This  proved to be an useful tool for detecting biomarkers which are associated with depression and could assist in future hypothesis generation.

Orabi etal ,2018 stated that Social media can be considered as another platform for detection of metal disorder as its becomes an integral part of our life. Bottleneck here is in adopting supervised machine learning approaches such as deep neural networks due to the presence of insufficient amounts of annotated training data.

Chung et al (    ) assessed participants below threshold for a diagnosis of a mental disorder during a clinician-administered diagnostic interview with a K-CESD-R Mobile app. Algorithm.

Bin dhim et al (2015) stated that smartphone apps was an effective screening tool for depression and could be delivered across a large number of countries. These Apps have the potential to play a significant role in disease screening, self-management, monitoring, and health education, particularly amongst younger adults. In another research (Bin dhim et al, 2016) he claimed that a mobile phone depression-screening app motivated some users to seek a depression diagnosis which was a very positive outcome.

Review of Mobile apps

 

In this section we aim at presenting an overview of all the depression screening tests that are available on the mobile applications. These contribution towards these applications has been done by clinical psychologist, neurologist, computer scientist and data scientist. The applications are mainly focussed on questionnaire-based ones but also includes which have a greater lifestyle and medical benefits as evident from the reviews. The functionalities of depression apps are also explained in brief with their target user and benefits especially strengths and weakness in mind. At the end of this review a table summarises the salient features of this depression apps.

The following tools are fundamentally to assess the depression like symptoms

The Patient Health Questionnaire (PHQ-9), a 9-item, self-report questionnaire was used to assess the severity of depressive symptomatology within the previous two weeks(3). The PHQ-9 is one of the most reliable and validated measures of depressive symptoms. The following scores correlate with symptom severity: 0-4 none, 5-9 mild, 10-14 moderate, 15-19 moderately severe, 20-27 severe.

The Generalized Anxiety Disorder scale (GAD-7), a 7-item, brief self-report tool was used to assess the frequency and severity of anxious thoughts and behaviours over the past two weeks(4). Scores of 5, 10, and 15 are taken as the cut-off points for mild, moderate and severe anxiety, respectively.

The Social Phobia Inventory (SPIN), a 17-item, self-rating tool was used to assess the severity of social anxiety symptomatology over the past week(5). The following scores correlate with symptom severity: 0-19 none, 20-30 mild, 31-40 moderate, 41-50 severe, more than 50 very severe.The Wilcoxon signed-rank test, a non-parametric statistical test(6), was used to demonstrate that the average symptom reduction is statistically significant (p < 0.001) from the baseline measurement.

These results and the details about the methodology of this research will be available in a peer-reviewed scientific publication soon.

Common Depression  Mobile Applications 

Depression is also called Major depression disorder and its different from General anxiety disorder. The technology advancements has allowed Virtual Counselling Using Natural Language Processing big time.

Woebot This application was developed in San Francisco by a start-up Woebot which used chatbot platform of machine learning and natural language processing. This allowed the patients to manage their mood and alleviate depression. This can be accessed through the Messenger platform of Facebook. The algorithm here is trained on cognitive behaviour therapy (CBT) methods therefore it learns the emotions user and reckons activities in order to balance their mood. It stimulates the users with relevant questions in order to assess their mood. This algorithm was developed and based on the significant clinical trial results conducted in partnership with University of Stanford. Woebot users are aged between 18-28 and experience “significant reductions in anxiety and depression” compared with the control group. This was confirmed in an e-book that was published by National Institutes of Health (NIH).The participants used this app daily or about 85percent  of time for a period of two weeks. Usefulness was measured using PHQ-9 (standard patient health questionnaire for depression) with  scores ranging from 0-20 (no to severe symptoms).This chatbot had 50,000 users in its first week, with roughly 1 million messages received each week. This was launched in June 2017  and is currently free which may not be the case as it progresses sustainable business model. This chatbot is based on CBT model which allows you to learn about yourself. It equips you with the tools which are tailormade to your moods and needs at that time. The users of this app can track their emotions through friendly conversation, can get guided for a range of problems through scientifically proven techniques, get help with managing their mood, relationship problems, grief, habits and addictions, the conversations can get personalized and can teach mindfulness.

Wysa This app is a competitor chatbot of Woebot and uses machine learning algorithms. The emotions of the users are learned through machine learning and natural language processing and an emotional balance is maintained by regular involvement. This involvement of the app is in response to the texts by the users and suggests appropriate cognitive-behavioural techniques (CBT), meditation, yoga and deep breathing techniques. This app has more than 200,000 users from 30 countries. Unfortunately, there are no clinical data or case studies on their website. This app is free, but access to human Wysa mental health professionals has a cost of $29 per month. It’s a package of mood tracker and booster, mindfulness coach and anxiety assistant. Its always available on demand to track your mood with friendly chats and helps fight stress and anxiety. Building in resilience and improving emotional health is the key to track happiness and mood. This app keeps your identity anonymous and all conversations are privacy protected. Wysa claims that it builds confidence, reduces self-doubt, manages anger, manages anxiety and deals with worry.

Pacifica This mobile application is designed by psychologist and gives tools based on Cognitive Behavioural Therapy, mindfulness meditation, relaxation, and mood/health tracking. A constant ongoing cycle of negative thoughts causes stress, anxiety, and depression. Thoughts cause physical feelings and emotions which cause actions. Pacifica is based on breaking this cycle using tools that target each of its components. Its has relaxation & mindfulness meditation audio tools which calms down in moments of stress or anxiety using one of Pacifica’s 25+ audio exercises. They have other deep breathing, muscle relaxation, positive visualization, mindfulness meditations etc techniques to calm down. These activities are set to relaxing soundscapes like ocean waves or thunderstorms. There are self-help guided paths which are designed by the psychologist which includes audio lessons and helpful activities to help you with stress, anxiety, or depression. These include an introductory Path, two Paths focused on CBT, and a mindfulness Path. The mood tracker would rate the mood, feelings throughout the day and add specific notes hence a pattern can be generated for situations or triggers which can be improved over time. A method of thought recording, journaling & analysis is also used as a tool by  Pacifica’s psychologist to learn how distorted thinking patterns contribute to your anxiety. The other tools are daily challenges & goal tracking, health tracker, student appointment and homework reminders, communities & groups which assist the users to face and overcome daily challenges in a stepwise method. Its free to download and use. Pacifica Full Access is based on auto-renewing subscription model and allows for unlimited use at a cost of $8.99  per month or $53.99 annually in USA.
Sunrise Health This application has emerged from another start-up called Johns Hopkins Technology Ventures. Its based in Baltimore and uses AI and predictive analytics. It monitors the patient activity and prevents the onset of a mental health crisis using support group texting. The algorithms are integrated in the app and are trained with data generated from a group users of  texts which are generally 5 to 7 in a group. Patterns in patient activity are recognized through predictive analytics, progress is measured to identify changes which could be indicative of a potential issue. Therapist are human moderators, are alerted immediately if the system predicts a potential issue with a patient. Besides anonymous support group texting there is an option of voice calls or face-to-face meetings with a therapist for the users of the app. Unfortunately, the advantage or case studies on this app reducing depression instances is not yet available.

Ginger.io  Ginger.io is a Californian based start-up and was founded in 2011.It uses machine learning to provide mental health support to users. Ginger.io has developed the algorithms by using training data from over 1 million consumers and through partnerships of over 40 healthcare company. This application is strikingly different to other app developers of similar applications as it provide users with access to a team of human mental health professionals and machine learning. It uses machine learning to learn from patient data and aligns care with patient goals and objectives. Members of the Ginger.io care team include emotional support coaches, licensed therapists and board-certified psychiatrists.

Ginger.io offers online, around-the-clock emotional support. Coaches are available to chat anytime, at a moment’s notice—whether it’s the middle of the night, a weekend or a holiday. This app is free to employees and members of select organizations and companies in the US, UK and Canada. If your employer, school or organization offers Ginger.io, you can download the app to get started today.
This app has Unlimited Chat and anytime with emotional support coach and receive replies in minutes, Video Sessions to meet virtually  licensed therapist or board certified psychiatrist, personalized Self-care to receive clinically-validated, self-care content that’s tailored to your individual goals and needs, confidential and Secure where any information shared is protected by HIPAA and is completely confidential. The app is downloaded, matched with a coach immediately and  its ready to go. The age must be above 18 years to use this app. This app developers has raised $28.2 million and lead investors include Kaiser Permanente Investors and Khosla Ventures. They have subscription packages from $129-$349 per month. Unfortunately, case studies on the impact of Ginger.io on patients and healthcare systems are not currently available on the company’s website.

Mindstrong Health This company was founded by Californian based startup in 2014 and uses machine learning to  diagnose and treat behavioural health disorders by interpreting data generated from mobile users. It uses data mining and focusses primarily on “digital phenotyping.” Digital phenotyping is a method of quantifying individual characteristics by analysing data generated from an individual’s use of smartphones and other personal digital devices (Harvard researchers,2016). They have trained its machine learning algorithms on an equivalent of 200 person-years of cognitive data. This is equivalent to the combined measurement of individuals and their time contribution – from three clinical studies. Predictive analytics then interpret the data and recommends correlations between specific digital activities and brain activity. The aim here is to “define signals correlated with cognition, brain imaging and mood in patients with depression.” (NIH-Stanford clinical trial).This is a research based organisation with main focus on viability of digital phenotyping as proven by its series of clinical trials. They have been able to raise $14 million, reportedly.

MOODPATH This app is an interactive depression and anxiety screening program. In this app the users psychological, emotional and physical wellbeing are tracked and a personalized mental health assessment is generated which could be discussed with a therapists. The app includes two-week depression screening, assessment of mental wellbeing, understand the mood psychology, use of Moodpath- mood journal/mood diary on smartphone, learn about psychotherapy and mental health, talk to a health professional with the help of the assessment. It helps to track, monitor, and understand complaints in a structured manner. During and at the end of the program helpful information on psychology, signs of depression, therapy and mental health followed by a detailed summary which could be presented to healthcare professionals is possible. This is available in German and English.

Youper(You +Super) This AI Assistant is based on therapy and meditation. It engages in quick and insightful conversations and assist in overcoming ups and downs, stress, depression, and anxiety. The daily question of “How are you?” is responded by selecting one of many color-coded moods. Based on the response youper helps to improve your mood, overcome stress and symptoms of anxiety or depression. This app has been developed and led by psychiatrist Dr. Jose Hamilton. This app utilizes artificial intelligence and draws the essence from psychological therapies including Cognitive Behavioural Therapy (CBT), Acceptance and Commitment Therapy (ACT) and Meditation with a personalized technique. Youper’s colourful charts enables one to understand emotions and improve behaviour and builds a dynamic record of mind and emotional health. Youper is delving deep into
users emotions, thoughts and behaviours are how are they connected. The Emotional Health Atlas  research projects is generating insights into the most common emotions and the factors influencing moods and behaviours. Analysed data from 100,000 individuals indicates more than 80% of them improved their mood by talking to Youper. Youper claims that the average length of a conversation to produce positive changes is 7 minutes. The data shows that using Youper at least once a week significantly reduces symptoms of depression, anxiety, and social anxiety.

SPARX(Smart, Positive, Active, Realistic, X-factor thoughts) SPARX is an online e-therapy tool provided by the University of Auckland, and funded by the Prime Minister’s Youth Mental Health Project. SPARX is free in New Zealand.   SPARX has been proven to help young people with mild to moderate depression and anxiety. Sparx is an app supported by Mental health foundation and Ministry of health, NZ. It’s a computer program and a mobile app that helps young people with mild to moderate depression, feeling anxious or stressed. They have a Mood Quiz which guides whether SPARX is appropriate or not. Youth have developed this and is based on ‘talking therapy’ also called Cognitive Behavioural Therapy. CBT teaches skills to cope with negative thoughts and feelings. CBT helps in balanced thinking and doing things that gives a sense of achievement. The app is based on self-help game avatar journey where you will meet different characters, solve puzzles and complete mini games. There are seven levels and each level takes about half an hour. Users aim to complete 2 levels per week. Its targeted towards young people who are feeling down, depressed or anxious. This study had been conducted in NZ and the results were published in the British Medical Journal in 2012. Three doctoral projects evaluated SPARX with Maori, Rainbow or same/both sex attracted youth and young people in Alternative Education. SPARX was effective for youths 12 to 19 years old seeking help for depression; reduced depression, anxiety, feelings of hopelessness and improved quality of life; These changes lasted for at least three months; worked better for those with more depression (but still within mild-moderate range);worked equally well across different ethnic groups in New Zealand; worked equally well for girls and boys and older and younger youths; worked equally well across the age group of teenagers 12 to 19 years; appeared to work better when users completed at least half of the modules (i.e. at least four levels). Sparx created by Associate Professor Sally Merry, Dr Karolina Stasiak, Dr Theresa (Terry) Fleming, Dr Matt Shepherd and Dr Mathijs Lucassen.Associate Professor Sally Merry is a Child and Adolescent Psychiatrist, Head of Department of Psychological Medicine and Director of The Werry Centre for Child and Adolescent Mental Health.Dr Stasiak also coordinated the main study of SPARX.Drs Fleming, Shepherd and Lucassen also carried out doctoral studies of SPARX.Metia Interactive developed the game. The National Institute for Health Innovation (NIHI) at University of Auckland hosts and supports SPARX online.  Sparx is high privacy encryption and firewall technology. At the end of the program seven levels a check is made on how the users feels and to monitor if SPARX is effective; to communicate with you about our goods and services from time to time; to generate an overview of the user base; to analyse trends and demographics to assist us to improve our services; facilitate our internal business operations, including to fulfil our legal requirements, quality assurance and management purposes.

Patient Health Questionnaire: PHQ-9 is the depression module The PHQ-9 is a 9-question tool given to depression patients in primary care to screen for the presence and severity of depression. It’s a 9-question depression scale from the Patient Health Questionnaire (PHQ). The results of the PHQ-9 may be used to make a depression diagnosis according to DSM-IV criteria and takes less than 3 minutes to complete. The total of all 9 responses from the PHQ-9 aims to predict the presence and severity of depression. Primary care providers frequently use the PHQ-9 to screen for depression in patients.

MoodPanda This app is a supportive Mood Diary and tracks mood and get anonymous community support of around 100,000 members. This App for iPhones or android, or web based. MoodPanda allows to measure daily moods and track the scores over  a period of time. You can connect to Twitter or Facebook to share your scores and be part of the supportive Moody Pandas community. MoodPanda can be used to show your website/forum/group member’s happiness or your local regions happiness (through Mood Maps). This app is easy to update and has a simple happiness rating – up to date your mood as often as you wish and add a note giving a reason. It has an optional privacy controls if you want your mood diary kept private away from the support community. This can be used from  mobile devices and the web app by a simple login. This app tracks mood, view graphs of mood ,interacts with the MoodPanda community and gets support from the community if needed, it has privacy mode which enable moodposts on the website anonymous  and also has a view mood calendars.

III. critical analysis and Data Analysis

WORKING ON THIS

A.      Data Description and Preparation

B.      Machine Learning Techniques Used

C.     Experimental Setting

IV.     Conclusion

The review of the mobile apps using machine learning as mode for depression screening has clearly indicated a deep lying potential of smart phone app and the use of technology in screening, self monitoring and then motivating the highly depressed patients to seek treatment. An early detection is better for the treatment and cure of this mental health issue.

V.  References (WORKING ON THIS)

Patel, M. J., Khalaf, A., & Aizenstein, H. J. (2016). Studying depression using imaging and machine learning methods. NeuroImage : Clinical, 10, 115–123.

Passos, I. C., Mwangi, B., Cao, B., Hamilton, J. E., Wu, M.-J., Zhang, X. Y., … Soares, J. C. (2016). Identifying a clinical signature of suicidality among patients with mood disorders: a pilot study using a machine learning approach. Journal of Affective Disorders, 193, 109–116.

Hilbert, K., Lueken, U., Muehlhan, M., & Beesdo‐Baum, K. (2017). Separating generalized anxiety disorder from major depression using clinical, hormonal, and structural MRI data: A multimodal machine learning study. Brain and Behavior, 7(3), e00633.

Kessler, R. C., van Loo, H. M., Wardenaar, K. J., Bossarte, R. M., Brenner, L. A., Cai, T., … Zaslavsky, A. M. (2016). Testing a machine-learning algorithm to predict the persistence and severity of major depressive disorder from baseline self-reports. Molecular Psychiatry, 21(10), 1366–1371.

Dipnall, J. F., Pasco, J. A., Berk, M., Williams, L. J., Dodd, S., Jacka, F. N., & Meyer, D. (2016). Fusing Data Mining, Machine Learning and Traditional Statistics to Detect Biomarkers Associated with Depression. PLoS ONE, 11(2), e0148195.

Foland-Ross, L. C., Sacchet, M. D., Prasad, G., Gilbert, B., Thompson, P. M., & Gotlib, I. H. (2015). Cortical Thickness Predicts the First Onset of Major Depression in Adolescence. International Journal of Developmental Neuroscience : The Official Journal of the International Society for Developmental Neuroscience, 46, 125–131.

Patel, M. J., Khalaf, A., & Aizenstein, H. J. (2016). Studying depression using imaging and machine learning methods. NeuroImage : Clinical, 10, 115–123.

Grist, R., Porter, J., & Stallard, P. (2017). Mental Health Mobile Apps for Preadolescents and Adolescents: A Systematic Review. Journal of medical Internet research, 19(5)

S.C. Wade, Benjamin & Joshi, Shantanu & Pirnia, Tara & Leaver, Amber & P. Woods, Roger & Thompson, Paul & Espinoza, Randall & Narr, Katherine. (2015). Random Forest Classification of Depression Status Based On Subcortical Brain Morphometry Following Electroconvulsive Therapy. Proceedings / IEEE International Symposium on Biomedical Imaging: from nano to macro. IEEE International Symposium on Biomedical Imaging. 2015. 92-96. 10.1109/ISBI.2015.7163824.

Merry Sally N, Stasiak Karolina, Shepherd Matthew, Frampton Chris, Fleming Theresa, Lucassen Mathijs F G et al.??
2.The effectiveness of SPARX, a computerised self help intervention for adolescents seeking help for depression: randomised controlled non-inferiority trial
BMJ2012; 344 :e2598

Fitzpatrick KK, Darcy A, Vierhile M
Delivering Cognitive Behavior Therapy to Young Adults With Symptoms of Depression and Anxiety Using a Fully Automated Conversational Agent (Woebot): A Randomized Controlled Trial
JMIR Ment Health 2017;4(2):e19

WHO Depression and Other Common Mental Disorders

Global Health Estimates, 2017.2

2017 IEEE 13th International Symposium on Autonomous Decentralized Systems Predicting Depression Levels Using Social Media Posts,Maryam Mohammed Aldarwish

Hafiz Farooq Ahmed

Mental Health America Online depression screening,2017 This report was researched, written and prepared by Theresa Nguyen, Michele Hellebuyck, Madeline Halpern, Maddy Reinert, and Danielle Fritze.

Proceedings of the Fifth Workshop on Computational Linguistics and Clinical Psychology: From Keyboard to Clinic, pages 88–97

New Orleans, Louisiana, June 5, 2018. c 2018 Association for Computational Linguistics Deep Learning for Depression Detection of Twitter Users

Ahmed Husseini Orabi, Prasadith Buddhitha, Mahmoud Husseini Orabi, Diana Inkpen

Teen posed as doctor at West Palm Beach hospital: police. (2015, January 16). Retrieved from http://www.nbcmiami.com/news/local/Teen-Posed-as-Doctor-at-West-Palm-Beach-Hospital-Police-288810831.html

Chung K, Jeon M-J, Park J, Lee S, Kim CO, Park JY (2018) Development and evaluation of a mobile-optimized daily self-rating depression screening app: A preliminary study. PLoS ONE 13(6): e0199118. https://doi.org/10.1371/journal pone.0199118

BinDhim NF, et al. J Am Med Inform Assoc 2015;22:29–34. doi:10.1136/amiajnl-2014-002840, Brief Communication

J Med Internet Res 2016 | vol. 18 | iss. 6 | e156 | p.1

POPULATING THIS

