Explainable ML models for a deeper insight on treatment decision for localized prostate cancer
Abstract Although there are several decision aids for the treatment of localized prostate cancer (PCa), there are limitations in the consistency and certainty of the information provided. We aimed to better understand the treatment decision process and develop a decision-predicting model considering...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-07-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-38162-1 |
| _version_ | 1797774210855075840 |
|---|---|
| author | Jang Hee Han Sungyup Lee Byounghwa Lee Ock-kee Baek Samuel L. Washington Annika Herlemann Peter E. Lonergan Peter R. Carroll Chang Wook Jeong Matthew R. Cooperberg |
| author_facet | Jang Hee Han Sungyup Lee Byounghwa Lee Ock-kee Baek Samuel L. Washington Annika Herlemann Peter E. Lonergan Peter R. Carroll Chang Wook Jeong Matthew R. Cooperberg |
| author_sort | Jang Hee Han |
| collection | DOAJ |
| description | Abstract Although there are several decision aids for the treatment of localized prostate cancer (PCa), there are limitations in the consistency and certainty of the information provided. We aimed to better understand the treatment decision process and develop a decision-predicting model considering oncologic, demographic, socioeconomic, and geographic factors. Men newly diagnosed with localized PCa between 2010 and 2015 from the Surveillance, Epidemiology, and End Results Prostate with Watchful Waiting database were included (n = 255,837). We designed two prediction models: (1) Active surveillance/watchful waiting (AS/WW), radical prostatectomy (RP), and radiation therapy (RT) decision prediction in the entire cohort. (2) Prediction of AS/WW decisions in the low-risk cohort. The discrimination of the model was evaluated using the multiclass area under the curve (AUC). A plausible Shapley additive explanations value was used to explain the model’s prediction results. Oncological variables affected the RP decisions most, whereas RT was highly affected by geographic factors. The dependence plot depicted the feature interactions in reaching a treatment decision. The decision predicting model achieved an overall multiclass AUC of 0.77, whereas 0.74 was confirmed for the low-risk model. Using a large population-based real-world database, we unraveled the complex decision-making process and visualized nonlinear feature interactions in localized PCa. |
| first_indexed | 2024-03-12T22:17:43Z |
| format | Article |
| id | doaj.art-4578e0b672504e5ea491806cad64b0a3 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-12T22:17:43Z |
| publishDate | 2023-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-4578e0b672504e5ea491806cad64b0a32023-07-23T11:13:40ZengNature PortfolioScientific Reports2045-23222023-07-011311810.1038/s41598-023-38162-1Explainable ML models for a deeper insight on treatment decision for localized prostate cancerJang Hee Han0Sungyup Lee1Byounghwa Lee2Ock-kee Baek3Samuel L. Washington4Annika Herlemann5Peter E. Lonergan6Peter R. Carroll7Chang Wook Jeong8Matthew R. Cooperberg9Department of Urology, Seoul National University HospitalElectronics and Telecommunications Research Institute (ETRI)Electronics and Telecommunications Research Institute (ETRI)Electronics and Telecommunications Research Institute (ETRI)Department of Urology, Helen Diller Family Comprehensive Cancer Center, University of CaliforniaDepartment of Urology, Helen Diller Family Comprehensive Cancer Center, University of CaliforniaDepartment of Urology, Helen Diller Family Comprehensive Cancer Center, University of CaliforniaDepartment of Urology, Helen Diller Family Comprehensive Cancer Center, University of CaliforniaDepartment of Urology, Seoul National University HospitalDepartment of Urology, Helen Diller Family Comprehensive Cancer Center, University of CaliforniaAbstract Although there are several decision aids for the treatment of localized prostate cancer (PCa), there are limitations in the consistency and certainty of the information provided. We aimed to better understand the treatment decision process and develop a decision-predicting model considering oncologic, demographic, socioeconomic, and geographic factors. Men newly diagnosed with localized PCa between 2010 and 2015 from the Surveillance, Epidemiology, and End Results Prostate with Watchful Waiting database were included (n = 255,837). We designed two prediction models: (1) Active surveillance/watchful waiting (AS/WW), radical prostatectomy (RP), and radiation therapy (RT) decision prediction in the entire cohort. (2) Prediction of AS/WW decisions in the low-risk cohort. The discrimination of the model was evaluated using the multiclass area under the curve (AUC). A plausible Shapley additive explanations value was used to explain the model’s prediction results. Oncological variables affected the RP decisions most, whereas RT was highly affected by geographic factors. The dependence plot depicted the feature interactions in reaching a treatment decision. The decision predicting model achieved an overall multiclass AUC of 0.77, whereas 0.74 was confirmed for the low-risk model. Using a large population-based real-world database, we unraveled the complex decision-making process and visualized nonlinear feature interactions in localized PCa.https://doi.org/10.1038/s41598-023-38162-1 |
| spellingShingle | Jang Hee Han Sungyup Lee Byounghwa Lee Ock-kee Baek Samuel L. Washington Annika Herlemann Peter E. Lonergan Peter R. Carroll Chang Wook Jeong Matthew R. Cooperberg Explainable ML models for a deeper insight on treatment decision for localized prostate cancer Scientific Reports |
| title | Explainable ML models for a deeper insight on treatment decision for localized prostate cancer |
| title_full | Explainable ML models for a deeper insight on treatment decision for localized prostate cancer |
| title_fullStr | Explainable ML models for a deeper insight on treatment decision for localized prostate cancer |
| title_full_unstemmed | Explainable ML models for a deeper insight on treatment decision for localized prostate cancer |
| title_short | Explainable ML models for a deeper insight on treatment decision for localized prostate cancer |
| title_sort | explainable ml models for a deeper insight on treatment decision for localized prostate cancer |
| url | https://doi.org/10.1038/s41598-023-38162-1 |
| work_keys_str_mv | AT jangheehan explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT sungyuplee explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT byounghwalee explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT ockkeebaek explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT samuellwashington explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT annikaherlemann explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT peterelonergan explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT peterrcarroll explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT changwookjeong explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer AT matthewrcooperberg explainablemlmodelsforadeeperinsightontreatmentdecisionforlocalizedprostatecancer |