The project for objective measures using computational psychiatry technology (PROMPT): Rationale, design, and
Taishiro Kishimoto1, Akihiro Takamiya1, Kuo-Ching Liang1
1Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku, Tokyo, 160-8582, Japan.
Contemporary Clinical Trials Communications
|September 11, 2020
Summary
This study aims to develop objective digital biomarkers for assessing depressive and neurocognitive disorders. Computational psychiatry and machine learning will analyze speech, motion, and activity data to improve diagnosis and treatment monitoring.
Area of Science:
- Computational psychiatry
- Digital biomarkers
- Machine learning in mental health
Background:
- Depressive and neurocognitive disorders are leading causes of global disability.
- Lack of objective biomarkers hinders treatment assessment and drug development.
- New technologies enable quantification of clinically relevant behavioral features.
Purpose of the Study:
- To develop objective, noninvasive, and easy-to-use biomarkers for assessing depressive and neurocognitive disorders.
- To guide clinical decision-making and reduce clinical trial failure rates.
- To leverage computational psychiatry and machine learning for objective mental health assessment.
Main Methods:
- Recruitment of patients with major depressive disorder, bipolar disorder, neurocognitive disorders, and healthy controls.
- Conducting 10-minute conversational interviews recorded via RGB/infrared cameras and microphones.
- Utilizing wearable devices and advanced software for data processing and machine learning analysis.
Main Results:
- Machine learning models are employed to predict symptom presence, severity, and changes over time.
- Analysis of multimodal data including video, audio, and wearable sensor information.
- Focus on extracting features that reflect disorder severity across diverse patient samples.
Conclusions:
- The Project for Objective Measures Using Computational Psychiatry Technology (PROMPT) aims to create practical biomarkers.
- Objective measures can improve clinical practice and streamline psychiatric drug development.
- Addressing sample variability is crucial for robust biomarker extraction.
Keywords:
AMED, Japan Agency for Medical Research and DevelopmentAdabag, Adaptive BaggingAdaboost, Adaptive BoostingBD, Bipolar disorderBDI-II, Beck Depression Inventory, Second EditionBNN, Bayesian Neural NetworksCDR, Clinical Dementia RatingCDT, Clock Drawing TestCNN, Convolutional Neural NetworksCPP, cepstral peak prominenceDSM-5, Diagnostic and Statistical Manual of Mental Disorders, Fifth EditionDepressionF0, fundamental frequencyF1, F2, F3, first, second, and third formant frequenciesFedRAMP, Federal Risk and Authorization Management ProgramGCNN, Gated Convolutional Neural NetworksGDS, Geriatric Depression ScaleHAM-D, Hamilton Depression Rating ScaleIEC, International Electrotechnical CommissionISO, International Organization for StandardizationLM, Wechsler Memory Scale-Revised Logical MemoryLSTM, Long Short-Term Memory NetworksM.I.N.I., Mini-International Neuropsychiatric InterviewMADRS, Montgomery-Asberg Depression Rating ScaleMARS, Motor Agitation and Retardation ScaleMCI, mild cognitive impairmentMDD, Major depressive disorderMFCC, mel-frequency cepstrum coefficientsMMSE, Mini-Mental State ExaminationMRI, magnetic resonance imagingMachine learningMoCA, Montreal Cognitive AssessmentNPI, Neuropsychiatric InventoryNatural language processingNeurocognitive disorderPET, positron emission tomographyPROMPT, Project for Objective Measures Using Computational Psychiatry TechnologyPSQI, Pittsburgh Sleep Quality IndexRF, Random ForestRGB, red, green, blueSCID, Structural Clinical Interview for DSM-5SVM, Support Vector MachineSVR, Support Vector RegressionScreeningUI, uncertainty intervalUMIN, University Hospital Medical Information NetworkUV, ultravioletYLDs, years lived with disabilityYMRS, Young Mania Rating ScaleMore Related Videos
13:09Using Brain Activation nir-HEG/Q-EEG and Execution Measures CPTs in a ADHD Assessment Protocol
Published on: April 1, 2018
10.7K
16:23Automated, Quantitative Cognitive/Behavioral Screening of Mice: For Genetics, Pharmacology, Animal Cognition and Undergraduate Instruction
Published on: February 26, 2014
14.7K