Mohammadhadi Khorrami1, Kaustav Bera1, Patrick Leo1, Pranjal Vaidya1, Pradnya Patil2, Rajat Thawani3, Priya Velu4, Prabhakar Rajiah5, Mehdi Alilou1, Humberto Choi6, Michael D Feldman4, Robert C Gilkeson7, Philip Linden8, Pingfu Fu9, Harvey Pass10, Vamsidhar Velcheti10, Anant Madabhushi11. 1. Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, USA. 2. Department of Solid Tumor Oncology, Cleveland Clinic, Cleveland, OH, USA. 3. Department of Internal Medicine, Maimonides Medical Center, Brooklyn, NY, USA. 4. Pathology and Laboratory Medicine, Hospital of the University of Pennsylvania, USA. 5. Department of Radiology, UT Southwestern Medical Center, Dallas, TX, USA. 6. Department of Pulmonary Medicine, Cleveland Clinic, Cleveland, OH, USA. 7. Department of Radiology, University Hospitals of Cleveland, OH, USA. 8. Thoracic and Esophageal Surgery Department, University Hospitals of Cleveland, OH, USA. 9. Department of Population and Quantitative Health Sciences, CWRU, Cleveland, OH, USA. 10. Department of Hematology and Oncology, NYU Langone Health, New York, NY, USA. 11. Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, USA. Electronic address: anant.madabhushi@case.edu.
Abstract
OBJECTIVES: To evaluate whether combining stability and discriminability criteria in building radiomic classifiers will improve the prognosis of cancer recurrence in early stage non-small cell lung cancer on non-contrast computer tomography (CT). MATERIALS AND METHODS: CT scans of 610 patients with early stage (IA, IB, IIA) NSCLC from four independent cohorts were evaluated. A total of 350 patients from Cleveland Clinic Foundation and University of Pennsylvania were divided into two equal sets for training (D1) and validation set (D2). 80 patients from The Cancer Genome Atlas Lung Adenocarcinoma and Squamous Cell Carcinoma and 195 patients from The Cancer Imaging Archive, were used as independent second (D3) and third (D4) validation sets. A linear discriminant analysis (LDA) classifier was built based on the most stable and discriminate features. In addition, a radiomic risk score (RRS) was generated by using least absolute shrinkage and selection operator, Cox regression model to predict time to progression (TTP) following surgery. RESULTS: A feature selection strategy focusing on both feature discriminability and stability resulted in the classifier having a higher discriminability on validation datasets compared to the discriminability alone criteria in discriminating cancer recurrence (D2, AUC of 0.75 vs. 0.65; D3, 0.74 vs. 0.62; D4, 0.76 vs. 0.63). The RRS generated by most stable-discriminating features was significantly associated with TTP compared to discriminating alone criteria (HR = 1.66, C-index of 0.72 vs. HR = 1.04, C-index of 0.62). CONCLUSION: Accounting for both stability and discriminability yielded a more generalizable classifier for predicting cancer recurrence and TTP in early stage NSCLC.
OBJECTIVES: To evaluate whether combining stability and discriminability criteria in building radiomic classifiers will improve the prognosis of cancer recurrence in early stage non-small cell lung cancer on non-contrast computer tomography (CT). MATERIALS AND METHODS: CT scans of 610 patients with early stage (IA, IB, IIA) NSCLC from four independent cohorts were evaluated. A total of 350 patients from Cleveland Clinic Foundation and University of Pennsylvania were divided into two equal sets for training (D1) and validation set (D2). 80 patients from The Cancer Genome Atlas Lung Adenocarcinoma and Squamous Cell Carcinoma and 195 patients from The Cancer Imaging Archive, were used as independent second (D3) and third (D4) validation sets. A linear discriminant analysis (LDA) classifier was built based on the most stable and discriminate features. In addition, a radiomic risk score (RRS) was generated by using least absolute shrinkage and selection operator, Cox regression model to predict time to progression (TTP) following surgery. RESULTS: A feature selection strategy focusing on both feature discriminability and stability resulted in the classifier having a higher discriminability on validation datasets compared to the discriminability alone criteria in discriminating cancer recurrence (D2, AUC of 0.75 vs. 0.65; D3, 0.74 vs. 0.62; D4, 0.76 vs. 0.63). The RRS generated by most stable-discriminating features was significantly associated with TTP compared to discriminating alone criteria (HR = 1.66, C-index of 0.72 vs. HR = 1.04, C-index of 0.62). CONCLUSION: Accounting for both stability and discriminability yielded a more generalizable classifier for predicting cancer recurrence and TTP in early stage NSCLC.
Authors: Berend C Stoel; M Els Bakker; Jan Stolk; Asger Dirksen; Robert A Stockley; Eeva Piitulainen; Erich W Russi; Johan H C Reiber Journal: Invest Radiol Date: 2004-01 Impact factor: 6.016
Authors: Patrick Leo; George Lee; Natalie N C Shih; Robin Elliott; Michael D Feldman; Anant Madabhushi Journal: J Med Imaging (Bellingham) Date: 2016-10-24
Authors: Ruben T H M Larue; Janna E van Timmeren; Evelyn E C de Jong; Giacomo Feliciani; Ralph T H Leijenaar; Wendy M J Schreurs; Meindert N Sosef; Frank H P J Raat; Frans H R van der Zande; Marco Das; Wouter van Elmpt; Philippe Lambin Journal: Acta Oncol Date: 2017-09-08 Impact factor: 4.089
Authors: Kenneth Clark; Bruce Vendt; Kirk Smith; John Freymann; Justin Kirby; Paul Koppel; Stephen Moore; Stanley Phillips; David Maffitt; Michael Pringle; Lawrence Tarbox; Fred Prior Journal: J Digit Imaging Date: 2013-12 Impact factor: 4.056
Authors: Timothy Winton; Robert Livingston; David Johnson; James Rigas; Michael Johnston; Charles Butts; Yvon Cormier; Glenwood Goss; Richard Inculet; Eric Vallieres; Willard Fry; Drew Bethune; Joseph Ayoub; Keyue Ding; Lesley Seymour; Barbara Graham; Ming-Sound Tsao; David Gandara; Kenneth Kesler; Todd Demmy; Frances Shepherd Journal: N Engl J Med Date: 2005-06-23 Impact factor: 91.245
Authors: John R Zech; Marcus A Badgeley; Manway Liu; Anthony B Costa; Joseph J Titano; Eric Karl Oermann Journal: PLoS Med Date: 2018-11-06 Impact factor: 11.069
Authors: Michael A Labarbera; Thomas Atta-Fosu; Albert K Feeny; Marjan Firouznia; Meghan Mchale; Catherine Cantlay; Tyler Roach; Alexis Axtell; Paul Schoenhagen; John Barnard; Jonathan D Smith; David R Van Wagoner; Anant Madabhushi; Mina K Chung Journal: IEEE J Transl Eng Health Med Date: 2021-12-09
Authors: Mohammadhadi Khorrami; Kaustav Bera; Rajat Thawani; Prabhakar Rajiah; Amit Gupta; Pingfu Fu; Philip Linden; Nathan Pennell; Frank Jacono; Robert C Gilkeson; Vamsidhar Velcheti; Anant Madabhushi Journal: Eur J Cancer Date: 2021-03-17 Impact factor: 9.162
Authors: Arthur Jochems; Turkey Refaee; Henry C Woodruff; Philippe Lambin; Guangyao Wu; Abdalla Ibrahim; Chenggong Yan; Sebastian Sanduleanu Journal: Eur J Nucl Med Mol Imaging Date: 2021-03-11 Impact factor: 10.057