Zeynettin Akkus

Meredith VandeHaar, M. Clinch, I. Yilmaz et al.

Quality assurance is a critical but underexplored area in digital pathology, where even minor artifacts can have significant effects. Artifacts have been shown to negatively impact the performance of AI diagnostic models. In current practice, trained staff manually review digitized images prior to release of these slides to pathologists which are then used to render a diagnosis. Conventional image processing approaches, provide a foundation for detecting artifacts on digital pathology slides. However, current tools do not leverage deep learning, which has the potential to improve detection accuracy and scalability. Despite these advancements, methods for quality assurance in digital pathology remain limited, presenting a gap for innovation. We propose an AI algorithm designed to screen digital pathology slides by analyzing tiles and categorizing them into one of 10 predefined artifact types or as background. This algorithm identifies and localizes artifacts, creating a map that highlights regions of interest. By directing human operators to specific tiles affected by artifacts, the algorithm minimizes the time and effort required to manually review entire slides for quality issues. From internal archives and The Cancer Genome Atlas, 133 whole slide images were selected and 10 artifacts were annotated using an internally developed software ZAPP (Mayo Clinic, Jacksonville, FL). Ablation study of multiple models at different tile sizes and magnification was performed. InceptionResNet was selected. Single artifact models were trained and tested, followed by a limited multiple instance model with artifacts that performed well together (chatter, fold, and pen). From the results of this study we suggest a hybrid design for artifact screening composed of both single artifact binary models as well as multiple instance models to optimize detection of each artifact.

Mohamed Y. Elwazir, Zeynettin Akkus, Didem Oguz et al.

Echocardiography (echo) is an indispensable tool in a cardiologist's diagnostic armamentarium. To date, almost all echocardiographic parameters require time-consuming manual labeling and measurements by an experienced echocardiographer and exhibit significant variability, owing to the noisy and artifact-laden nature of echo images. For example, mitral inflow (MI) Doppler is used to assess left ventricular (LV) diastolic function, which is of paramount clinical importance to distinguish between different cardiac diseases. In the current work we present a fully automated workflow which leverages deep learning to a) label MI Doppler images acquired in an echo study, b) detect the envelope of MI Doppler signal, c) extract early and late filing (E and A wave) flow velocities and E-wave deceleration time from the envelope. We trained a variety of convolutional neural networks (CNN) models on 5544 images of 140 patients for predicting 24 image classes including MI Doppler images and obtained overall accuracy of 0.97 on 1737 images of 40 patients. Automated E and A wave velocity showed excellent correlation (Pearson R 0.99 and 0.98 respectively) and Bland Altman agreement (mean difference 0.06 and 0.05 m/s respectively and SD 0.03 for both) with the operator measurements. Deceleration time also showed good but lower correlation (Pearson R 0.82) and Bland-Altman agreement (mean difference: 34.1ms, SD: 30.9ms). These results demonstrate feasibility of Doppler echocardiography measurement automation and the promise of a fully automated echocardiography measurement package.

Tomas Sakinis, Fausto Milletari, Holger Roth et al.

Image segmentation plays an essential role in medicine for both diagnostic and interventional tasks. Segmentation approaches are either manual, semi-automated or fully-automated. Manual segmentation offers full control over the quality of the results, but is tedious, time consuming and prone to operator bias. Fully automated methods require no human effort, but often deliver sub-optimal results without providing users with the means to make corrections. Semi-automated approaches keep users in control of the results by providing means for interaction, but the main challenge is to offer a good trade-off between precision and required interaction. In this paper we present a deep learning (DL) based semi-automated segmentation approach that aims to be a "smart" interactive tool for region of interest delineation in medical images. We demonstrate its use for segmenting multiple organs on computed tomography (CT) of the abdomen. Our approach solves some of the most pressing clinical challenges: (i) it requires only one to a few user clicks to deliver excellent 2D segmentations in a fast and reliable fashion; (ii) it can generalize to previously unseen structures and "corner cases"; (iii) it delivers results that can be corrected quickly in a smart and intuitive way up to an arbitrary degree of precision chosen by the user and (iv) ensures high accuracy. We present our approach and compare it to other techniques and previous work to show the advantages brought by our method.

Zeynettin Akkus, Issa Ali, Jiri Sedlar et al.

Objective: Several studies have associated codeletion of chromosome arms 1p/19q in low-grade gliomas (LGG) with positive response to treatment and longer progression free survival. Therefore, predicting 1p/19q status is crucial for effective treatment planning of LGG. In this study, we predict the 1p/19q status from MR images using convolutional neural networks (CNN), which could be a noninvasive alternative to surgical biopsy and histopathological analysis. Method: Our method consists of three main steps: image registration, tumor segmentation, and classification of 1p/19q status using CNN. We included a total of 159 LGG with 3 image slices each who had biopsy-proven 1p/19q status (57 nondeleted and 102 codeleted) and preoperative postcontrast-T1 (T1C) and T2 images. We divided our data into training, validation, and test sets. The training data was balanced for equal class probability and then augmented with iterations of random translational shift, rotation, and horizontal and vertical flips to increase the size of the training set. We shuffled and augmented the training data to counter overfitting in each epoch. Finally, we evaluated several configurations of a multi-scale CNN architecture until training and validation accuracies became consistent. Results: The results of the best performing configuration on the unseen test set were 93.3% (sensitivity), 82.22% (specificity), and 87.7% (accuracy). Conclusion: Multi-scale CNN with their self-learning capability provides promising results for predicting 1p/19q status noninvasively based on T1C and T2 images. Significance: Predicting 1p/19q status noninvasively from MR images would allow selecting effective treatment strategies for LGG patients without the need for surgical biopsy.