F Christiansen1, E L Epstein1, E Smedberg2, M Åkerlund3, K Smith4, E Epstein2. 1. School of Engineering Sciences, KTH Royal Institute of Technology, Stockholm, Sweden. 2. Department of Clinical Science and Education, Karolinska Institutet, and Department of Obstetrics and Gynecology, Södersjukhuset, Stockholm, Sweden. 3. Harvard Extension School, Harvard University, Cambridge, MA, USA. 4. Science for Life Laboratory, School of Electrical Engineering and Computer Science, KTH Royal Institute of Technology, Stockholm, Sweden.
Abstract
OBJECTIVES: To develop and test the performance of computerized ultrasound image analysis using deep neural networks (DNNs) in discriminating between benign and malignant ovarian tumors and to compare its diagnostic accuracy with that of subjective assessment (SA) by an ultrasound expert. METHODS: We included 3077 (grayscale, n = 1927; power Doppler, n = 1150) ultrasound images from 758 women with ovarian tumors, who were classified prospectively by expert ultrasound examiners according to IOTA (International Ovarian Tumor Analysis) terms and definitions. Histological outcome from surgery (n = 634) or long-term (≥ 3 years) follow-up (n = 124) served as the gold standard. The dataset was split into a training set (n = 508; 314 benign and 194 malignant), a validation set (n = 100; 60 benign and 40 malignant) and a test set (n = 150; 75 benign and 75 malignant). We used transfer learning on three pre-trained DNNs: VGG16, ResNet50 and MobileNet. Each model was trained, and the outputs calibrated, using temperature scaling. An ensemble of the three models was then used to estimate the probability of malignancy based on all images from a given case. The DNN ensemble classified the tumors as benign or malignant (Ovry-Dx1 model); or as benign, inconclusive or malignant (Ovry-Dx2 model). The diagnostic performance of the DNN models, in terms of sensitivity and specificity, was compared to that of SA for classifying ovarian tumors in the test set. RESULTS: At a sensitivity of 96.0%, Ovry-Dx1 had a specificity similar to that of SA (86.7% vs 88.0%; P = 1.0). Ovry-Dx2 had a sensitivity of 97.1% and a specificity of 93.7%, when designating 12.7% of the lesions as inconclusive. By complimenting Ovry-Dx2 with SA in inconclusive cases, the overall sensitivity (96.0%) and specificity (89.3%) were not significantly different from using SA in all cases (P = 1.0). CONCLUSION: Ultrasound image analysis using DNNs can predict ovarian malignancy with a diagnostic accuracy comparable to that of human expert examiners, indicating that these models may have a role in the triage of women with an ovarian tumor.
OBJECTIVES: To develop and test the performance of computerized ultrasound image analysis using deep neural networks (DNNs) in discriminating between benign and malignant ovarian tumors and to compare its diagnostic accuracy with that of subjective assessment (SA) by an ultrasound expert. METHODS: We included 3077 (grayscale, n = 1927; power Doppler, n = 1150) ultrasound images from 758 women with ovarian tumors, who were classified prospectively by expert ultrasound examiners according to IOTA (International Ovarian Tumor Analysis) terms and definitions. Histological outcome from surgery (n = 634) or long-term (≥ 3 years) follow-up (n = 124) served as the gold standard. The dataset was split into a training set (n = 508; 314 benign and 194 malignant), a validation set (n = 100; 60 benign and 40 malignant) and a test set (n = 150; 75 benign and 75 malignant). We used transfer learning on three pre-trained DNNs: VGG16, ResNet50 and MobileNet. Each model was trained, and the outputs calibrated, using temperature scaling. An ensemble of the three models was then used to estimate the probability of malignancy based on all images from a given case. The DNN ensemble classified the tumors as benign or malignant (Ovry-Dx1 model); or as benign, inconclusive or malignant (Ovry-Dx2 model). The diagnostic performance of the DNN models, in terms of sensitivity and specificity, was compared to that of SA for classifying ovarian tumors in the test set. RESULTS: At a sensitivity of 96.0%, Ovry-Dx1 had a specificity similar to that of SA (86.7% vs 88.0%; P = 1.0). Ovry-Dx2 had a sensitivity of 97.1% and a specificity of 93.7%, when designating 12.7% of the lesions as inconclusive. By complimenting Ovry-Dx2 with SA in inconclusive cases, the overall sensitivity (96.0%) and specificity (89.3%) were not significantly different from using SA in all cases (P = 1.0). CONCLUSION: Ultrasound image analysis using DNNs can predict ovarian malignancy with a diagnostic accuracy comparable to that of human expert examiners, indicating that these models may have a role in the triage of women with an ovarian tumor.
Authors: Caroline Van Holsbeke; Anneleen Daemen; Joseph Yazbek; Tom K Holland; Tom Bourne; Tinne Mesens; Lore Lannoo; Anne-Sophie Boes; Annelies Joos; Arne Van De Vijver; Nele Roggen; Bart de Moor; Eric de Jonge; Antonia C Testa; Lil Valentin; Davor Jurkovic; Dirk Timmerman Journal: Gynecol Obstet Invest Date: 2009-12-11 Impact factor: 2.031
Authors: Wouter Froyman; Chiara Landolfo; Bavo De Cock; Laure Wynants; Povilas Sladkevicius; Antonia Carla Testa; Caroline Van Holsbeke; Ekaterini Domali; Robert Fruscio; Elisabeth Epstein; Maria José Dos Santos Bernardo; Dorella Franchi; Marek Jerzy Kudla; Valentina Chiappa; Juan Luis Alcazar; Francesco Paolo Giuseppe Leone; Francesca Buonomo; Lauri Hochberg; Maria Elisabetta Coccia; Stefano Guerriero; Nandita Deo; Ligita Jokubkiene; Jeroen Kaijser; An Coosemans; Ignace Vergote; Jan Yvan Verbakel; Tom Bourne; Ben Van Calster; Lil Valentin; Dirk Timmerman Journal: Lancet Oncol Date: 2019-02-05 Impact factor: 41.316
Authors: Dirk Timmerman; Lieveke Ameye; Daniela Fischerova; Elisabeth Epstein; Gian Benedetto Melis; Stefano Guerriero; Caroline Van Holsbeke; Luca Savelli; Robert Fruscio; Andrea Alberto Lissoni; Antonia Carla Testa; Joan Veldman; Ignace Vergote; Sabine Van Huffel; Tom Bourne; Lil Valentin Journal: BMJ Date: 2010-12-14
Authors: Scott Mayer McKinney; Marcin Sieniek; Varun Godbole; Jonathan Godwin; Natasha Antropova; Hutan Ashrafian; Trevor Back; Mary Chesus; Greg S Corrado; Ara Darzi; Mozziyar Etemadi; Florencia Garcia-Vicente; Fiona J Gilbert; Mark Halling-Brown; Demis Hassabis; Sunny Jansen; Alan Karthikesalingam; Christopher J Kelly; Dominic King; Joseph R Ledsam; David Melnick; Hormuz Mostofi; Lily Peng; Joshua Jay Reicher; Bernardino Romera-Paredes; Richard Sidebottom; Mustafa Suleyman; Daniel Tse; Kenneth C Young; Jeffrey De Fauw; Shravya Shetty Journal: Nature Date: 2020-01-01 Impact factor: 49.962
Authors: Eiron John Lugtu; Denise Bernadette Ramos; Alliah Jen Agpalza; Erika Antoinette Cabral; Rian Paolo Carandang; Jennica Elia Dee; Angelica Martinez; Julius Eleazar Jose; Abegail Santillan; Ruth Bangaoil; Pia Marie Albano; Rock Christian Tomas Journal: PLoS One Date: 2022-05-12 Impact factor: 3.752