生存予測のためのPROの効果的な使用:NSCLC患者のトランスフォーマーベースのモデリング
D Dudas1, T J Dilling2, H Jim3
1LMU Hospital, Department of Radiation Oncology, Munich, Germany; H. Lee Moffitt Cancer Center and Research Institute, Department of Machine Learning, Tampa, FL, USA.
まとめ
患者によって報告されたアウトカム (PROs) を使用したトランスフォーマーモデルは,SBRTで治療された初期段階の非小細胞肺がん (NSCLC) 患者の生存予測の正確性を改善しました. 食欲低下と痛みは,生存の最も有意な予測要因でした.
科学分野:
- 腫瘍学 腫瘍学
- データサイエンス データサイエンス
- 医療情報工学 医療情報工学
背景:
- 正確な生存率の予測は,がん学における患者中心のケア,治療計画,緩和ケア紹介において極めて重要です.
- 現在の臨床生存推定値は過度に楽観的で,患者の生活の質 (QoL) を低下させる可能性があります.
- 患者によって報告されたアウトカム (PROs) は,診断の正確性を高める貴重な予測指標です.
研究 の 目的:
- 早期非小細胞肺癌 (NSCLC) 患者における生存予測の精度を向上させるために,縦方向のPRO軌道を活用するためのトランスフォーマーアーキテクチャを探求する.
- トランスフォーマーモデルを使用して生存予測のための最も予後的に関連するPRO症状を特定する.
主な方法:
- SBRTで治療された475人の初期段階のNSCLC患者の縦断PROデータ (エドマントン症状評価スケール - ESAS) を分析するために,トランスフォーマーベースのモデルが開発されました.
- モデルには,全生存率 (OS) の予測のためのPROs,臨床,および人口統計的変数が含まれ,コックス比例危険性回帰および共同確率モデルに対するパフォーマンスを比較しました.
- SHAPLEY ADDITIVE EXPLANATION (SHAP) 値と,c-インデックスとAUCによって導かれた後退排除は,モデル解釈と症状識別に使用されました.
主要な成果:
- トランスフォーマーモデルは,発見セットでは0.753のc-indexと0.862のAUCのクロス検証を達成し,ホールドアウトテストセットでは0.694のc-indexと0.785のAUCを達成しました.
- トランスフォーマーモデルは,従来のコックスと共同確率生存モデルを大幅に上回った.
- SHAPの分析では,食欲の減少,痛み,全体的な幸福感,呼吸困難が生存予測の最も重要な予後症状として特定されました.
結論:
- 縦線PROを統合したトランスフォーマーベースの生存モデルは,SBRTで治療されたNSCLC患者の予後精度を大幅に改善します.
- 食欲と痛みの喪失は,生存の最も強力な予測要因として特定され,その次に総合的な幸福感と呼吸困難が続きます.
- 標的に,症状に焦点を当てたPROトラッキングは,生存率の推定を向上させ,通常の腫瘍学ケアにおける臨床実施を合理化することができます.
関連する概念動画
01:33Bacterial Transformation
60.2K
In 1928, bacteriologist Frederick Griffith worked on a vaccine for pneumonia, which is caused by Streptococcus pneumoniae bacteria. Griffith studied two pneumonia strains in mice: one pathogenic and one non-pathogenic. Only the pathogenic strain killed host mice.
Griffith made an unexpected discovery when he killed the pathogenic strain and mixed its remains with the live, non-pathogenic strain. Not only did the mixture kill host mice, but it also contained living pathogenic bacteria that...
Griffith made an unexpected discovery when he killed the pathogenic strain and mixed its remains with the live, non-pathogenic strain. Not only did the mixture kill host mice, but it also contained living pathogenic bacteria that...
60.2K
01:27Pharmacodynamics in Geriatric Patients: Effects of Age
242
Age-related pharmacokinetic changes are extensively documented, but understanding age-related pharmacodynamic alterations is relatively limited. This knowledge gap can be partly attributed to the complexity of developing appropriate measures of drug responses compared to bioanalytical methods for determining drug concentrations.Most information regarding age-related differences in human pharmacodynamics originates from cross-sectional studies. However, these studies assume that observed mean...
242
02:27Predicting Molecular Geometry
46.2K
VSEPR Theory for Determination of Electron Pair Geometries
46.2K
01:18Survival Curves
749
Survival curves are graphical representations that depict the survival experience of a population over time, offering an intuitive way to track the proportion of individuals who remain event-free at each time point. These curves are widely used in fields such as medicine, public health, and reliability engineering to visualize and compare survival probabilities across different groups or conditions.
The Kaplan-Meier estimator is the most common method for constructing survival curves. This...
The Kaplan-Meier estimator is the most common method for constructing survival curves. This...
749
01:19Survival Tree
437
Survival trees are a non-parametric method used in survival analysis to model the relationship between a set of covariates and the time until an event of interest occurs, often referred to as the "time-to-event" or "survival time." This method is particularly useful when dealing with censored data, where the event has not occurred for some individuals by the end of the study period, or when the exact time of the event is unknown.
Building a Survival Tree
Constructing a...
Building a Survival Tree
Constructing a...
437
01:03Prediction Intervals
3.4K
The interval estimate of any variable is known as the prediction interval. It helps decide if a point estimate is dependable.
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.
3.4K