Predicting conversion to psychosis using machine learning: response to Cannon
BackgroundWe previously reported that machine learning could be used to predict conversion to psychosis in individuals at clinical high risk (CHR) for psychosis with up to 90% accuracy using the North American Prodrome Longitudinal Study-3 (NAPLS-3) dataset. A definitive test of our predictive model...
Saved in:
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-01-01
|
| Series: | Frontiers in Psychiatry |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fpsyt.2024.1520173/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841527595215093760 |
|---|---|
| author | Jason Smucny Tyrone D. Cannon Tyrone D. Cannon Carrie E. Bearden Carrie E. Bearden Jean Addington Kristen S. Cadenhead Barbara A. Cornblatt Matcheri Keshavan Daniel H. Mathalon Diana O. Perkins William Stone Elaine F. Walker Scott W. Woods Scott W. Woods Ian Davidson Cameron S. Carter |
| author_facet | Jason Smucny Tyrone D. Cannon Tyrone D. Cannon Carrie E. Bearden Carrie E. Bearden Jean Addington Kristen S. Cadenhead Barbara A. Cornblatt Matcheri Keshavan Daniel H. Mathalon Diana O. Perkins William Stone Elaine F. Walker Scott W. Woods Scott W. Woods Ian Davidson Cameron S. Carter |
| author_sort | Jason Smucny |
| collection | DOAJ |
| description | BackgroundWe previously reported that machine learning could be used to predict conversion to psychosis in individuals at clinical high risk (CHR) for psychosis with up to 90% accuracy using the North American Prodrome Longitudinal Study-3 (NAPLS-3) dataset. A definitive test of our predictive model that was trained on the NAPLS-3 data, however, requires further support through implementation in an independent dataset. In this report we tested for model generalization using the previous iteration of NAPLS-3, the NAPLS-2, using the identical machine learning algorithms employed in our previous study.MethodStandard machine learning algorithms were trained to predict conversion to psychosis in clinical high risk individuals on the NAPLS-3 dataset and tested on the NAPLS-2 dataset.ResultsNAPLS-2 and -3 individuals significantly differed on most features used in machine learning models. All models performed above chance, with Naive Bayes and random forest methods showing the best overall performance. Importantly, however, overall performance did not match those previously observed when using only NAPLS-3 data.ConclusionThe results of this study suggest that a machine learning model trained to predict conversion to psychosis on one dataset can be used to train an independent dataset. Performance on the test set was not in the range necessary for clinical application, however. Possible reasons that limited performance are discussed. |
| format | Article |
| id | doaj-art-37f1364b29e84db68874f32d9a100ae8 |
| institution | Kabale University |
| issn | 1664-0640 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Psychiatry |
| spelling | doaj-art-37f1364b29e84db68874f32d9a100ae82025-01-15T11:24:35ZengFrontiers Media S.A.Frontiers in Psychiatry1664-06402025-01-011510.3389/fpsyt.2024.15201731520173Predicting conversion to psychosis using machine learning: response to CannonJason Smucny0Tyrone D. Cannon1Tyrone D. Cannon2Carrie E. Bearden3Carrie E. Bearden4Jean Addington5Kristen S. Cadenhead6Barbara A. Cornblatt7Matcheri Keshavan8Daniel H. Mathalon9Diana O. Perkins10William Stone11Elaine F. Walker12Scott W. Woods13Scott W. Woods14Ian Davidson15Cameron S. Carter16Department of Psychiatry, University of California, Davis, Davis, CA, United StatesDepartment of Psychology, Yale University, New Haven, CT, United StatesDepartment of Psychiatry, Yale University, New Haven, CT, United StatesDepartment of Psychiatry, Semel Institute for Neuroscience and Human Behavior, University of California, Los Angeles, Los Angeles, CA, United StatesBiobehavioral Sciences and Psychology, Semel Institute for Neuroscience and Human Behavior, University of California, Los Angeles, Los Angeles, CA, United StatesDepartment of Psychiatry, University of Calgary, Calgary, AB, CanadaDepartment of Psychiatry, University of North Carolina Chapel Hill, Chapel Hill, NC, United StatesDepartment of Psychiatry Research, Zucker Hillside Hospital, New York, NY, United StatesDepartment of Psychiatry, Harvard University, Cambridge, MA, United States0Department of Psychiatry, University of California, San Francisco, San Francisco, CA, United States1Department of Psychiatry, University of San Diego, San Diego, CA, United StatesDepartment of Psychiatry, Harvard University, Cambridge, MA, United States2Department of Psychiatry, Emory University, Atlanta, GA, United StatesDepartment of Psychology, Yale University, New Haven, CT, United StatesDepartment of Psychiatry, Yale University, New Haven, CT, United States3Department of Computer Science, University of California, Davis, Davis, CA, United States4Department of Psychiatry, University of California, Irvine, Irvine, CA, United StatesBackgroundWe previously reported that machine learning could be used to predict conversion to psychosis in individuals at clinical high risk (CHR) for psychosis with up to 90% accuracy using the North American Prodrome Longitudinal Study-3 (NAPLS-3) dataset. A definitive test of our predictive model that was trained on the NAPLS-3 data, however, requires further support through implementation in an independent dataset. In this report we tested for model generalization using the previous iteration of NAPLS-3, the NAPLS-2, using the identical machine learning algorithms employed in our previous study.MethodStandard machine learning algorithms were trained to predict conversion to psychosis in clinical high risk individuals on the NAPLS-3 dataset and tested on the NAPLS-2 dataset.ResultsNAPLS-2 and -3 individuals significantly differed on most features used in machine learning models. All models performed above chance, with Naive Bayes and random forest methods showing the best overall performance. Importantly, however, overall performance did not match those previously observed when using only NAPLS-3 data.ConclusionThe results of this study suggest that a machine learning model trained to predict conversion to psychosis on one dataset can be used to train an independent dataset. Performance on the test set was not in the range necessary for clinical application, however. Possible reasons that limited performance are discussed.https://www.frontiersin.org/articles/10.3389/fpsyt.2024.1520173/fullschizophreniaclinical high risk (CHR)NAPLSout of sample evaluationscale of psychosis risk symptomsgeneralizability |
| spellingShingle | Jason Smucny Tyrone D. Cannon Tyrone D. Cannon Carrie E. Bearden Carrie E. Bearden Jean Addington Kristen S. Cadenhead Barbara A. Cornblatt Matcheri Keshavan Daniel H. Mathalon Diana O. Perkins William Stone Elaine F. Walker Scott W. Woods Scott W. Woods Ian Davidson Cameron S. Carter Predicting conversion to psychosis using machine learning: response to Cannon Frontiers in Psychiatry schizophrenia clinical high risk (CHR) NAPLS out of sample evaluation scale of psychosis risk symptoms generalizability |
| title | Predicting conversion to psychosis using machine learning: response to Cannon |
| title_full | Predicting conversion to psychosis using machine learning: response to Cannon |
| title_fullStr | Predicting conversion to psychosis using machine learning: response to Cannon |
| title_full_unstemmed | Predicting conversion to psychosis using machine learning: response to Cannon |
| title_short | Predicting conversion to psychosis using machine learning: response to Cannon |
| title_sort | predicting conversion to psychosis using machine learning response to cannon |
| topic | schizophrenia clinical high risk (CHR) NAPLS out of sample evaluation scale of psychosis risk symptoms generalizability |
| url | https://www.frontiersin.org/articles/10.3389/fpsyt.2024.1520173/full |
| work_keys_str_mv | AT jasonsmucny predictingconversiontopsychosisusingmachinelearningresponsetocannon AT tyronedcannon predictingconversiontopsychosisusingmachinelearningresponsetocannon AT tyronedcannon predictingconversiontopsychosisusingmachinelearningresponsetocannon AT carrieebearden predictingconversiontopsychosisusingmachinelearningresponsetocannon AT carrieebearden predictingconversiontopsychosisusingmachinelearningresponsetocannon AT jeanaddington predictingconversiontopsychosisusingmachinelearningresponsetocannon AT kristenscadenhead predictingconversiontopsychosisusingmachinelearningresponsetocannon AT barbaraacornblatt predictingconversiontopsychosisusingmachinelearningresponsetocannon AT matcherikeshavan predictingconversiontopsychosisusingmachinelearningresponsetocannon AT danielhmathalon predictingconversiontopsychosisusingmachinelearningresponsetocannon AT dianaoperkins predictingconversiontopsychosisusingmachinelearningresponsetocannon AT williamstone predictingconversiontopsychosisusingmachinelearningresponsetocannon AT elainefwalker predictingconversiontopsychosisusingmachinelearningresponsetocannon AT scottwwoods predictingconversiontopsychosisusingmachinelearningresponsetocannon AT scottwwoods predictingconversiontopsychosisusingmachinelearningresponsetocannon AT iandavidson predictingconversiontopsychosisusingmachinelearningresponsetocannon AT cameronscarter predictingconversiontopsychosisusingmachinelearningresponsetocannon |