ABSTRACT
Objective: To evaluate the performance of an artificial intelligence (AI)-based medical analysis tool called uAI-HematomaCare in automatically detecting, classifying, and measuring the volume of acute spontaneous intracranial hemorrhage (ICH) on non-contrast-enhanced CT (NCCT) images.
Methods: A 1:2 matched case-control study design was used. The analysis included NCCT images from 70 cases with ICH and 140 non-ICH cases (70 cases with hemorrhage-like hyperdensities). Initially, the algorithm's performance in detecting and classifying acute ICH subtypes was compared to that of two radiologists. The criterion standard was determined by consensus of two blinded neuroradiology experts. Subsequently, the algorithm and two radiologists quantified the volumes of intraparenchymal hemorrhage (IPH), intraventricular hemorrhage (IVH), subarachnoid hemorrhage (SAH), and total ICH using manual segmentation. The performance of the measurement methods was evaluated by calculating the intraclass correlation coefficient (ICC) and absolute error.
Results: The algorithm showed a sensitivity of 100% and specificity of 80% for detecting any ICH, while radiologist 1 showed a sensitivity of 100% and specificity of 99.29%, and radiologist 2 showed a sensitivity of 100% and specificity of 100%. Regarding the classification of ICH subtypes, the algorithm showed a higher sensitivity in detecting IVH (100%) and SAH (100%), but lower specificity (97.71% and 95.19%, respectively) compared to the two radiologists. ICCs ranging from 0.932 to 0.983 indicated good agreement between the algorithm and manual segmentation methods for measuring summation and individual hemorrhage volumes (IPH, IVH, and SAH). Bland-Altman analysis of absolute errors revealed that the algorithm performed well in quantifying the volume of total ICH and the three subtypes, but radiologist 1 had better overall performance.
Conclusion: The AI-based algorithm analysis tool, uAI-HematomaCare, demonstrated a high accuracy in automatically detecting, classifying, and quantifying acute spontaneous ICH. However, further improvements are needed to enhance its ability to differentiate between hyperdensities and ICH.
- Roh D J, Asonye I S, Carvalho Poyraz F, Magid-Bernstein J R, Joiner E F, Avadhani R, Awad I. 2022. Intraventricular Hemorrhage Expansion in the Clear Iii Trial: A Post Hoc Exploratory Analysis. Stroke, 53, 6 (Jun 2022), 1847-1853. https://doi.org/10.1161/strokeaha.121.037438Google ScholarCross Ref
- Steiner T, Juvela S, Unterberg A, Jung C, Forsting M, Rinkel G. 2013. European Stroke Organization Guidelines for the Management of Intracranial Aneurysms and Subarachnoid Haemorrhage. Cerebrovascular diseases (Basel, Switzerland), 35, 2 (Dec 2013), 93-112. https://doi.org/10.1159/000346087Google ScholarCross Ref
- Glover M t, Almeida R R, Schaefer P W, Lev M H, Mehan W A, Jr. 2017. Quantifying the Impact of Noninterpretive Tasks on Radiology Report Turn-around Times. Journal of the American College of Radiology : JACR, 14, 11 (Nov 2017), 1498-1503. https://doi.org/10.1016/j.jacr.2017.07.023Google ScholarCross Ref
- Gibson E, Georgescu B, Ceccaldi P, Trigan P H, Yoo Y, Das J, Re T J. 2022. Artificial Intelligence with Statistical Confidence Scores for Detection of Acute or Subacute Hemorrhage on Noncontrast Ct Head Scans. Radiol Artif Intell, 4, 3 (May 2022), e210115. https://doi.org/10.1148/ryai.210115Google ScholarCross Ref
- Zhao X, Chen K, Wu G, Zhang G, Zhou X, Lv C, Wu S. 2021. Deep Learning Shows Good Reliability for Automatic Segmentation and Volume Measurement of Brain Hemorrhage, Intraventricular Extension, and Peripheral Edema. European radiology, 31, 7 (Jul 2021), 5012-5020. https://doi.org/10.1007/s00330-020-07558-2Google ScholarCross Ref
- Gou X, He X. 2021. Deep Learning-Based Detection and Diagnosis of Subarachnoid Hemorrhage. J Healthc Eng, 2021, 2021), 9639419. https://doi.org/10.1155/2021/9639419Google ScholarCross Ref
- Chang P D, Kuoy E, Grinband J, Weinberg B D, Thompson M, Homo R, Chen J. 2018. Hybrid 3d/2d Convolutional Neural Network for Hemorrhage Evaluation on Head Ct. AJNR. American journal of neuroradiology, 39, 9 (Sep 2018), 1609-1616. https://doi.org/10.3174/ajnr.A5742Google ScholarCross Ref
- Schmitt N, Mokli Y, Weyland C S, Gerry S, Herweh C, Ringleb P A, Nagel S. 2022. Automated Detection and Segmentation of Intracranial Hemorrhage Suspect Hyperdensities in Non-Contrast-Enhanced Ct Scans of Acute Stroke Patients. European radiology, 32, 4 (Apr 2022), 2246-2254. https://doi.org/10.1007/s00330-021-08352-4Google ScholarCross Ref
- Seu T. RSNA 2019 Intracranial Hemorrhage Detection: Code for 1st Place Solution in Intracranial Hemorrhage Detection Challenge. Retrieved May 27, 2023 from https://github.com/SeuTao/RSNA2019_1st_place_solution.Google Scholar
- Lee H, Yune S, Mansouri M, Kim M, Tajmir S H, Guerrier C E, Ebert S A. 2019. An Explainable Deep-Learning Algorithm for the Detection of Acute Intracranial Haemorrhage from Small Datasets. Nature biomedical engineering, 3, 3 (Mar 2019), 173-182. https://doi.org/10.1038/s41551-018-0324-9Google ScholarCross Ref
- Cho J, Park K S, Karki M, Lee E, Ko S, Kim J K, Lee D. 2019. Improving Sensitivity on Identification and Delineation of Intracranial Hemorrhage Lesion Using Cascaded Deep Learning Models. Journal of digital imaging, 32, 3 (Jun 2019), 450-461. https://doi.org/10.1007/s10278-018-00172-1Google ScholarCross Ref
- Ye H, Gao F, Yin Y, Guo D, Zhao P, Lu Y, Wang X. 2019. Precise Diagnosis of Intracranial Hemorrhage and Subtypes Using a Three-Dimensional Joint Convolutional and Recurrent Neural Network. European radiology, 29, 11 (Nov 2019), 6191-6201. https://doi.org/10.1007/s00330-019-06163-2Google ScholarCross Ref
- Dhar R, Falcone G J, Chen Y, Hamzehloo A, Kirsch E P, Noche R B, Roth K. 2020. Deep Learning for Automated Measurement of Hemorrhage and Perihematomal Edema in Supratentorial Intracerebral Hemorrhage. Stroke, 51, 2 (Feb 2020), 648-651. https://doi.org/10.1161/STROKEAHA.119.027657Google ScholarCross Ref
- Ironside N, Chen C J, Mutasa S, Sim J L, Marfatia S, Roh D, Ding D. 2019. Fully Automated Segmentation Algorithm for Hematoma Volumetric Analysis in Spontaneous Intracerebral Hemorrhage. Stroke, 50, 12 (Dec 2019), 3416-3423. https://doi.org/10.1161/strokeaha.119.026561Google ScholarCross Ref
- Chung K J, Kuang H, Federico A, Choi H S, Kasickova L, Al Sultan A S, Horn M. 2021. Semi-Automatic Measurement of Intracranial Hemorrhage Growth on Non-Contrast Ct. International journal of stroke : official journal of the International Stroke Society, 16, 2 (Feb 2021), 192-199. https://doi.org/10.1177/1747493019895704Google ScholarCross Ref
Index Terms
- Artificial intelligence based algorithm for automatic diagnosis and quantification of acute spontaneous intracranial hemorrhage
Recommendations
Diabetic Macular Edema Screened by Handheld Smartphone-based Retinal Camera and Artificial Intelligence
AbstractOur aim was to assess the tomographic presence of diabetic macular edema in type 2 diabetes patients screened for diabetic retinopathy with color fundus photographs and an artificial intelligence algorithm. Color fundus photographs obtained with a ...
Automatic Detection of the Existence of Subarachnoid Hemorrhage from Clinical CT Images
Subarachnoid hemorrhage (SAH) is a medical emergency which can lead to death or severe disability. Misinterpretation of computed tomography (CT) in patients with SAH is a common problem. How to improve the accuracy of diagnosis is a great challenge to ...
Application of artificial intelligence techniques for non-alcoholic fatty liver disease diagnosis: A systematic review (2005–2023)
Highlights- NAFLD is a liver disease that is becoming more common throughout the world.
- This article provides a comprehensive review of the studies conducted in the area of utilizing artificial intelligence to determine the prevalence of NAFLD/...
Abstract Background and objectivesNon-alcoholic fatty liver disease (NAFLD) is a common liver disease with a rapidly growing incidence worldwide. For prognostication and therapeutic decisions, it is important to distinguish the pathological stages of ...
Comments