Learning deep neural networks' architectures using differential evolution. Case study: Medical imaging processing

Smaranda Belciug, Smaranda Belciug

Abstract

The COVID-19 pandemic has changed the way we practice medicine. Cancer patient and obstetric care landscapes have been distorted. Delaying cancer diagnosis or maternal-fetal monitoring increased the number of preventable deaths or pregnancy complications. One solution is using Artificial Intelligence to help the medical personnel establish the diagnosis in a faster and more accurate manner. Deep learning is the state-of-the-art solution for image classification. Researchers manually design the structure of fix deep learning neural networks structures and afterwards verify their performance. The goal of this paper is to propose a potential method for learning deep network architectures automatically. As the number of networks architectures increases exponentially with the number of convolutional layers in the network, we propose a differential evolution algorithm to traverse the search space. At first, we propose a way to encode the network structure as a candidate solution of fixed-length integer array, followed by the initialization of differential evolution method. A set of random individuals is generated, followed by mutation, recombination, and selection. At each generation the individuals with the poorest loss values are eliminated and replaced with more competitive individuals. The model has been tested on three cancer datasets containing MRI scans and histopathological images and two maternal-fetal screening ultrasound images. The novel proposed method has been compared and statistically benchmarked to four state-of-the-art deep learning networks: VGG16, ResNet50, Inception V3, and DenseNet169. The experimental results showed that the model is competitive to other state-of-the-art models, obtaining accuracies between 78.73% and 99.50% depending on the dataset it had been applied on.

Keywords: Cancer MRI scan; Cancer histopathological image; Deep learning; Differential evolution; Maternal-fetal ultrasound; Statistical assessment.

Conflict of interest statement

The authors declare that there is no conflict of interest.

Copyright © 2022 The Author. Published by Elsevier Ltd.. All rights reserved.

Figures

Fig. 1
Fig. 1
DE/CNN architecture.
Fig. 2
Fig. 2
(a) glioma tumor, (b) meningioma tumor, (c) pituitary tumor, (d) no tumor. (https://doi.org/10.34740/kaggle/dsv/1183165), [82].
Fig. 3
Fig. 3
(a) adenocarcinoma, (b) squamos cell carcinoma, (c) benign tissue (https://arxiv.org/abs/1912.12142v1, https://github.com/tampapath/lung_colon_image_set) [83].
Fig. 4
Fig. 4
(a) adenocarcinoma, (b) benign tissue (https://arxiv.org/abs/1912.12142v1, https://github.com/tampapath/lung_colon_image_set) [83].
Fig. 5
Fig. 5
(a) fetal abdomen, (b) fetal brain, (c) maternal cervix, (d) fetal femur, (e) fetal thorax, (f) other (https://zenodo.org/record/3904280#.YfjeTPVBzL9 [37],.
Fig. 6
Fig. 6
(a). trans-ventricular, (b) trans-thalamic, (c) trans-cerebellar (https://zenodo.org/record/3904280#.YfjeTPVBzL9 [37].
Fig. 7
Fig. 7
One-way ANOVA – Least Square Means (a) Bc dataset, (b) Lc dataset, (c), Cc dataset, (d) FP dataset, (e) FB dataset.
Fig. 8
Fig. 8
Distribution boxplot together with p-level: (a) Bc dataset, (b) Lc dataset, (c) Cc dataset, (d) Fp dataset, (e) Fb dataset.

References

    1. Feletto E., Grogan P., Nickson C., Smith M., Canfell K. How has COVID-19 impacted cancer screening? Adaptation of services and the future outlook in Australia. Publ. Health Res. Pract. 2020;30(4)
    1. Ng K.Y.Y., Zhou S., Tan S.H., Ishak N.D.B., Goh Z.Z.S., Chua Z.Y., et al. Understanding the psychological impact of COVID-19 pandemic on patients with cancer, their caregivers, and health care workers in Singapore. JCO Global Oncol. 2020;6:1494–1509.
    1. van de Haar J., Hoes L.R., Coles C.E., Seamon K., Frohling S., Jager D., et al. Caring for patients with cancer in the COVID-19 era. Nat. Med. 2020;26(5):665–671.
    1. van Dorn A. COVID-19 and readjusting clinical trials. Lancet (London, England) 2020;396(1025):523–524.
    1. Young A.M., Ashbury F.D., Schapira L., Scotte F., Ripamonti C.I., Olver I.N. Support Care Cancer; 2020. Uncertainty upon Uncertainty: Supportive Care for Cancer and COVID-19.
    1. Deprest J., Choolani M., Chervenak F., et al. Fetal diagnosis and therapy during the COVID-19 Pandemic: guidance on behalf of the international fetal medicine and surgery society. Fetal Diagn. Ther. 2020;47:689–698. doi: 10.1159/000508254.
    1. Mazur-Bialy A.I., Bogucka D.K., Tim S., Oplawski M. Pregnancy and Childbirth in the COVID-19 Era – the course of disease and maternal-fetal transmission. J. Clin. Med. 2020;9(11):3749. doi: 10.3390/jcm9113749.
    1. Chmielewska B., Barratt I., Townsend R., et al. Effects of the COVID-19 pandemic on maternal and perinatal outcomes: a systematic review and meta-analysis. Lancet Global Health. 2021 doi: 10.1016/S2214-109X(21)00079-6.
    1. Alkatout I., Biebl M., Momenimovahed Z., Giovannucci E., Hadavandsiri F., Salehiniya H., Allahgoli L. How COVID-19 affected cancer screening programs? A systematic review. Front. Oncol. 2021;11
    1. Gong K., Xu Z., Cai Z., Chen Y., Wang Z. Internet hospitals help prevent and control the epidemic of COVID-19 in China: multicenter user profiling study. J. Med. Internet Res. 2020;22(4)
    1. Cheng S.Y., Chen C.F., He H.C., Chang L.C., Hsu W.F., Wu M.S., et al. Impact of COVID-19 pandemic on fecal immunochemical test screening uptake and compliance to diagnostic colonoscopy. J. Gastroenterol. Hepatol. 2020;20 doi: 10.1111/jgh15325.
    1. Dinmohamed A.G., Visser O., Verhoeven R.H.A., Louwman M.W.J., van Nederveen F.H., Willems S.M., et al. Fewer Cancer diagnoses during the COVID-19 epidemic in The Netherlands. Lancet Oncol. 2020;21(6):750–751.
    1. Patt D., Gordan L., Diaz M., Okon T., Grady L., Harmison M., et al. Impact of COVID-19 on cancer care: how the pandemic is delaying cancer diagnosis and treatment for American seniors. JCO Clin. Cancer Inf. 2020;4:1059–1071.
    1. Lang M., Yeung T., Shepard J.O., Sharma A., Petranovic M., Flores E.J., et al. Operational Challenges of a low-dose CT lung cancer screening program during the coronavirus disease 2019 pandemic. Chest. 2020;159(3):1288–1291.
    1. Mathew R.V., Oliver K., Farrimond S., et al. Brain tumors and COVID-19: the patients and caregiver experience. Neurooncol. Adv. 2020;2(1) vdaa104.
    1. Dube R., Kar S.S. COVID-19 in pregnancy: the foetal perspective-a systematic review. Neonatology. 2020;4(1) doi: 10.1136/bmjpo-2020-000859.
    1. Salomon L., et al. A score-based method for quality control of fetal images at routine second trimester ultrasound examination. Prenat. Diagn. 2008;28(9):822–827.
    1. Paladini D. Sonography in obese and overweight pregnant women: clinical, medicolegal and tehncial issues. Ultrasound Obstet. Gynecol. 2009;33(6):720–729.
    1. Topol E.J. High performances medicine: the convergence of human and artificial intelligence. Nat. Med. 2019;25:44–46.
    1. Benjamens S., Dhunno P., Mesko B. The state of artificial intelligence-based FDA approved medical devices and algorithms: an online database. NPJ Digit. Med. 2020;3:118.
    1. Liu X., et al. A comparison of deep learning performances against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis. Lancet Digit. Health. 2019;1:e271–297.
    1. Kumar D., Wong A., Clausi D.A. Lung nodule classification using deep features in CT images. 12th Conf. Comput. Robot Vis. 2015:133–138.
    1. Sun W., Zheng B., Qian W. Computer aided lung cancer diagnosis with deep learning algorithms. Med. Imaging: Computer-Aided Diagnosis. 2016;9785
    1. Coudray N., Ocampo P.S., Sakellaropoulos T., et al. Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nat. Med. 2018;24(10):1559–1567.
    1. Song Q., Zhao L., Luo X., Dou X. Using deep learning for classification of lung nodules on computed tomography images. J. Healthc. Eng. 2017
    1. Bhatia S., Sinha Y., Goel L. Soft Comp for Probl Sol; Singer, Singapore: 2019. Lung Cancer Detection: a Deep Learning Approach; pp. 699–705.
    1. Teramoto A., Fujita H., Yamamuro O., Tamaki T. Automated detection of pulmonary nodules in PET/CT images: ensemble of false-positive reduction using a convolutional neural network technique. Med. Phys. 2016;43:2821–2827.
    1. Chen H., Zhao H., Shen J., Zhou R., Zhou Q. Supervised machine learning model for high dimensional gene data in colon cancer detection. IEEE Int. Congr. Big Data. 2015:134–141.
    1. Sirinukunwattana K., Raza S.E., Tsang Y.W., Snead D.R., Cree I.A., Rajpoot N.M. Locality sensitive deep learning for detection and classification of nuclei in routine colon cancer histology images. IEEE Trans. Med. Imag. 2016;35(5):1196–1206.
    1. Havaei M., Davy A., Warde-Farley D., Biard A., Courville A., Bengio Y., Pal C., Jodoin P.M., Larochelle H. 2015. Brain Tumor Segmentation with Deep Neural Networks. .
    1. Xiao Z., Huang R., Ding Y., Lan T., Dong R., Qin Z., Zhang X., Wang W. IEEE 6th International Conference on Computational Advances in Bio and Medical Sciences (ICCABS) 2016. A deep learning-based segmentation method for brain tumor in MR images; pp. 1–6.
    1. Dong H., Yang, et al. Automatic Brain Tumor Detection and Segmentation Using U-Net Based Fully Convolutional Networks, Annul Conference on Medical Image Understanding and Analysis. Springer; 2017. pp. 506–517.
    1. Rezaei M., et al. Springer; 2017. A Conditional Adversarial Network for Semantic Segmentation of Brain Tumor, International MICCAI Brainlesion Workshop; pp. 241–252.
    1. Zhao X., et al. A deep learning model integrating FCNNs and CRFs for brain tumor segmentation. Med. Image Anal. 2018;1(43):98–111.
    1. Munir K., Elahi H., Ayub A., Frezz F., Rizzi A. Cancer diagnosis using deep learning: a bibliographic review. Cancers. 2019;11(9):1235.
    1. Alom M.Z., et al. A state-of-the-art survey on deep learning theory and architectures. Elecronics. 2019;8(3):292.
    1. Burgos-Artizzu X.P., et al. vol. 19. Zenodo; 2020. FETAL_PLANES_DB: common maternal-fetal ultrasound images; p. 10200. (Nature Scientific Reports). 1.0.
    1. Matsuoka R., Komatsu M., et al. A novel deep learning based system for fetal cardiac screening. Ulstrasound Obstet. Gynecol. 2019 doi: 10.1002/uog.20945.
    1. Komatsu R., Matsuoka R., et al. Novel AI-guided ultrasound screening system for fetal heart can demonstrate finding in timeline diagram. Ultrasound Obstet. Gynecol. 2019 doi: 10.1002/uog.20796.
    1. Namburete A., et al. Fully automated alignment of 3D fetal brain ultrasound to a canonical reference space using multi-task learning. Med. Image Anal. 2018;46:1–14.
    1. Phillip M., et al. IEEE 16th International Symposium on Biomedical Imaging. 2019. Convolutional Neural Networks for automated fetal cardiac assessment using 4D B-Mode ultrasound; pp. 824–828.
    1. Torrents-Barrena J., et al. Assessment of radiomics and deep learning for the segmentation of fetal and maternal anatomy in magnetic resonance imaging and ultrasound. Acad. Radiol. 2019;S1076–6332(19):30575–30576.
    1. Wolpert D.H., Macready W.G. No free lunch theorems for optimization. IEEE Trans. Evol. Comput. 1997;1:67.
    1. Baker B., Gupta O., Naik N., Raskar R. International Conference on Learning Representations, ICLR; 2017. Designing neural network architectures using reinforcement learning; p. 2017.
    1. Cai H., Chen T., Zhang W., Yu Y., Wang J. Association for the Advancement of Artificial Intelligence; 2018. Efficient Architecture Search by Network Transformation; p. 2018.
    1. Zhong Z., Yan J., Liu C.L. vol. 2017. ICLR; 2018. (Practical Network Blocks Design with Q-Learning, International Conference on Learning Representations).
    1. Zoph B., Le Q.V. vol. 2017. ICLR; 2017. Neural architecture search with reinforcement learning. (International Conference on Learning Representations).
    1. Zoph B., Vasudevan V., Shlens J., Le Q.V. vol. 2018. 2018. (Learning Transferable Architectures for Scalable Image Recognition, Conference on Computer Vision and Pattern Recognition).
    1. Liu C., Zoph B., et al. European Conference on Computer Vision. 2018. Progressive neural architecture search; p. 2018.
    1. Miikkulainen R., Liang J.Z., et al. CoRR; 2017. Evolving Deep Neural Networks. abs/1703.00548.
    1. Real E., Moore S., et al. vol. 70. ICML; 2017. Large-scale evolution for image classifiers; pp. 2902–2911. (Proceedings of the 34th International Conference on Machine Learning). 17.
    1. Real E., Aggarwal A., Huang Y., Le Q.V. The Thirty-Third AAAI Conference on Artificial Intelligence, AAAI 2019, the Thirty-First Innovative Applications of Artificial Intelligence Conference, IAAI 2019, the Ninth AAAI Symposium on Educational Advances in Artificial Intelligence. EAAAI 2019; Honolulu, Hawaii, USA: 2019. Regularized evolution for image classifier architecture search; pp. 4780–4789.
    1. Sun Y., Xue B., Zhang M., Yen G.G. Evolving deep convolutional neural networks for image classification. IEEE Trans. Evol. Comput. 2020;24(2):394–407.
    1. Sun Y., Xue B., Zhang M., Yen G.G. Completely automated CNN architecture design based on blocks. IEEE Transact. Neural Networks Learn. Syst. 2020;31(4):1242–1254.
    1. Lindauer M., Hutter F. 2019. Best Practices for Scientific Research on Neural Architecture Search. .
    1. Williams R.J. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Mach. Learn. 1992;8:229–256.
    1. Schulman J., et al. 2017. Proximal Policy Optimization Algorithm. CoRR abs/1707.06347.
    1. Whitelam S., Tamblyn I. Learning to grow: control of material self-assembly using evolutionary reinforcement learning. Phys. Rev. E E. 2020;101 doi: 10.1103/PhysRevE.101.052604.
    1. Lomurno E., Samele S., Matteucci M., Ardagna D. GECCO’21: Proceedings of the Genetic and Evolutionary Computations Conference Companion. 2021. Pareto-optimal progressive neural architecture search; pp. 1726–1734.
    1. Stanley K.O. 2017. Neuroevolution: a Different Kind of Deep Learning.
    1. Liu H., Simonayan K., Vinyals O., Fernando C., Kavukcuoglu K. ICLR. 2018. Hierarchical representations for efficient architecture search.
    1. Stanley K.O., Miikkulainen R. Evolving neural networks through augmenting topologies. Evol. Comput. 2002;10:99–127.
    1. Hajewski J., Oliviera S., Xing X. 2020. Distributed Evolution of Deep Autoencoders. .
    1. Sun Y., Wang H., Xue B., Jin Y., Yen G.G., Zhang M. Surrogate-assisted evolutionary deep learning using an end-to-end random forest-based performance predictor. IEEE Trans. Evol. Comput. 2020;24(2):350–364.
    1. Whitelam S., Selin V., Park S.-W., Tamblyn I. Correspondence between neuroevolution and gradient descent. Nat. Commun. 2021;12:6317. doi: 10.1038/s41467-021-26568-2.
    1. Bahri Y., Kadmon J., Pennington J., Schoenholz S.S., Sohl-Dickstein J., Ganguli S. Statistical mechanics of deep learning. Annu. Rev. Condens. Matter Phys. 2020;11:501–528. doi: 10.1146/annurev-conmatphys-031119-050745.
    1. Stanley K.O., Clune J., Lehman J., Miikkulainen R. Designing neural networks through neuroevolution. Nat. Mach. Intell. 2019;1:24–35. doi: 10.1038/s42256-018-0006-z.
    1. Galvan E., Mooney P. Neuroevolution in deep neural networks: current trends and future challenges. IEEE Trans. Artif. Intell. 2021;2:476–493. doi: 10.1109/TAI.2021.3067574.
    1. Huang G., Sun Y., Liu Z., Sedra D., Weinberger K. European Conference on Compute Vision; 2016. Deep Networks with Stochastic Depth.
    1. Ioffe S., Szegedy C. International Conference on machine Learning; 2015. Accelerating Deep Network Training by Reducing Internal Covariate Shift; p. 2015.
    1. Storn R., Price K. Differential-evolution – a simple and efficient heuristic for global optimization over continuous spaces. J. Global Optim. 1997;11(4):341–359.
    1. Storn R., Price K. Differential evolution for multi-objective optimization. Evol. Comput. 2003;4:8–12. 2003.
    1. Omran M.G.H., Engelbrecht A.P. Self-adaptive differential evolution methods for unsupervised image classification. Proc. IEEE Conf. Cybern. Intell. Syst. 2006:1–6.
    1. Aslantas V., Tunckanat M. Proc. Of the IEEE Interantional Symposium on Intelligent Signal Processing (WISP) 2007. Differential evolution algorithm for segmentation of wound images.
    1. Dhahri H., Alimi A.M. The modified differential evolution and the RBF (MDE-RBF) neural network for time series prediction. Proc. Int. Joint Conf. Neural Network. 2006:2938–2943.
    1. Yang S., Gan Y.B., Qing A. Sideband suppression in time-modulated linear arrays by the differential evolution algorithm. IEEE Trans. Antenn. Propagations Lett. 2002;1(1):173–175.
    1. Kim H.K., Chong J.K., Park K.Y., Lowther D.A. Differential evolution strategy for constrained global optimization and application to practical engineering problems. IEEE Trans. Magn. 2007;43(4):1565–1568.
    1. Massa A., Pastorino M., Randazzo A. Optimization of the directivity of a monopulse antenna with a subarray weighting by a hybrid differential evolution method. IEEE Trans. Antenn. Propagations Lett. 2006;5(1):155–158.
    1. Su C.T., Lee C.S. Network reconfiguration of distribution systems using improved mixed-integer hybrid differential evolution. IEEE Trans. Power Deliv. 2003;18(3):1022–1027.
    1. Tasgetiren M.F., Suganthan P.N., Chua T.J., Al-Hajri A. Proceedings of the IEEE Congress on Evolutionary Computation (CEC ’09) 2009. Differential evolution algorithms for the generalized assignment problem; pp. 2606–2613.
    1. Sum-Im T., Taylor, Irving M.R., Song Y.H. Proceedings of the 42nd International Universities Power Engineering Conference (UPEC’07) 2007. A differential evolution algorithm for multistage transmission planning; pp. 357–364.
    1. Bhubaji S., Kadam A., Bhumkar P., Dedge S., Kanchan S. Brain tumor classification (MRI) Kaggle. 2020 doi: 10.34740/Kaggle/dsv/1183165.
    1. Borkowski A.A., Bui M.M., Thomas L.B., Wilson C.P., DeLand L.A., Mastorides S.M. 2019. Lung and Colon Cancer Histopathological Image Dataset (LC25000) arxiv:1912.12142v1 [eess.IV]
    1. Altman D.G. Chapman and Hall; New York: 1991. Practical Statistics for Medical Research.
    1. Belciug S. Elsevier; 2020. Artificial Intelligence in Cancer: Diagnostic to Tailored Treatment.
    1. Yap B.W., Sim C.H. Comparisons of various types of normality tests. J. Stat. Comput. Simulat. 2011;81(12):2141–2155. doi: 10.1080/00949655.2010.520163.
    1. Simonyan K., Zisserman A. 2014. Very Deep Convolutional Networks for Large-Scale Image Recognition. .
    1. He K., Zhang X., Ren S., Sun J. 2015. Deep Residual Learning for Image Recognition. .
    1. Szegedy C., Vanhoucke V., Ioffe S., Shlens J., Wojna Z. 2015. Rethinking the Inception Architecture for Computer Vision. . 3.
    1. Huang G., Liu Z., van der Maaten L., Weinberger K.Q. 2016. Densely Connected Convolutional Networks. .
    1. Demsar J. Statistical comparisons of classifiers over multiple data sets. J. Mach. Learn. Res. 2006;7:1–30.
    1. Seltman H. Experimental design and analysis. 2018.
    1. Kang J., Ullah Z., Gwak J. MRI-Based Brain Tumor Classification using ensemble of deep features and machine learning classifiers. Sensors. 2021;21(6):2222. doi: 10.3390/s21062222.
    1. Mangal S., Chaurasia A., Khajanchi A. 2020. Convolutional Neural Networks for Diagnosing Colon and Cancer Histopathological Images. arXiv:2009.03878.
    1. Hatuwal B.K., Thapa H.C. Lung cancer detection using convolutional neural network on histophatological images. Int. J. Comput. Trends Technol. 2020;68:21–24.
    1. Bukhari S.U.K., Asmara S., Bokhari S.K.A., et al. The histological diagnosis of colonic adenocarcinoma by applying partial self-supervised learning. medRxiv. 2020 doi: 10.1101/2020.08.15.20175760.
    1. Burgos-Artizzu X.P., Coronado-Guiterrez D., Valenzuela-Alcaraz B., et al. Evaluation of deep convolutional neural networks for automatic classification of common maternal fetal ultrasound planes. Sci. Rep. 2022;10:10200. doi: 10.10138/s41598-020-67076-5.

Source: PubMed

Sponsorer og samarbejdspartnere

Medicinske tilstande

Narkotikainterventioner

3
Abonner