The first case of glioma detected by an artificial intelligence algorithm running on real-time data in neurosurgery: illustrative case

Alperen Sozer Department of Neurosurgery

Search for other papers by Alperen Sozer in
Current site
jns
Google Scholar
PubMed
Close
 MD
,
Alp Ozgun Borcek Department of Neurosurgery, Division of Pediatric Neurosurgery, Gazi University Faculty of Medicine, and

Search for other papers by Alp Ozgun Borcek in
Current site
jns
Google Scholar
PubMed
Close
 MD
,
Seref Sagiroglu Department of Computer Engineering, Gazi University Faculty of Engineering, Ankara, Türkiye

Search for other papers by Seref Sagiroglu in
Current site
jns
Google Scholar
PubMed
Close
 PhD
,
Ali Poshtkouh Department of Neurosurgery

Search for other papers by Ali Poshtkouh in
Current site
jns
Google Scholar
PubMed
Close
 MSc
,
Zuhal Demirtas Department of Neurosurgery

Search for other papers by Zuhal Demirtas in
Current site
jns
Google Scholar
PubMed
Close
 MSc
,
Mehmet Melih Karaaslan Department of Neurosurgery

Search for other papers by Mehmet Melih Karaaslan in
Current site
jns
Google Scholar
PubMed
Close
 MD
,
Pelin Kuzucu Department of Neurosurgery

Search for other papers by Pelin Kuzucu in
Current site
jns
Google Scholar
PubMed
Close
 MD
, and
Emrah Celtikci Department of Neurosurgery

Search for other papers by Emrah Celtikci in
Current site
jns
Google Scholar
PubMed
Close
 MD
Open access

BACKGROUND

The aim of this paper is to report one of the significant applications of artificial intelligence (AI) and how it affects everyday clinical practice in neurosurgery. The authors present a case in which a patient was diagnosed via an AI algorithm during ongoing magnetic resonance imaging (MRI). According to this algorithm, the corresponding physicians were immediately warned, and the patient received prompt appropriate treatment.

OBSERVATIONS

A 46-year-old female presenting with nonspecific headache was admitted to undergo MRI. Scanning revealed an intraparenchymal mass that was detected by an AI algorithm running on real-time patient data while the patient was still in the MRI scanner. The day after MRI, a stereotactic biopsy was performed. The pathology report confirmed an isocitrate dehydrogenase wild-type diffuse glioma. The patient was referred to the oncology department for evaluation and immediate treatment.

LESSONS

This is the first report of a glioma diagnosed by an AI algorithm and a subsequent prompt operation in the literature—the first of many and an example of how AI will enhance clinical practice.

ABBREVIATIONS

AI = artificial intelligence; CT = computed tomography; MRI = magnetic resonance imaging

BACKGROUND

The aim of this paper is to report one of the significant applications of artificial intelligence (AI) and how it affects everyday clinical practice in neurosurgery. The authors present a case in which a patient was diagnosed via an AI algorithm during ongoing magnetic resonance imaging (MRI). According to this algorithm, the corresponding physicians were immediately warned, and the patient received prompt appropriate treatment.

OBSERVATIONS

A 46-year-old female presenting with nonspecific headache was admitted to undergo MRI. Scanning revealed an intraparenchymal mass that was detected by an AI algorithm running on real-time patient data while the patient was still in the MRI scanner. The day after MRI, a stereotactic biopsy was performed. The pathology report confirmed an isocitrate dehydrogenase wild-type diffuse glioma. The patient was referred to the oncology department for evaluation and immediate treatment.

LESSONS

This is the first report of a glioma diagnosed by an AI algorithm and a subsequent prompt operation in the literature—the first of many and an example of how AI will enhance clinical practice.

ABBREVIATIONS

AI = artificial intelligence; CT = computed tomography; MRI = magnetic resonance imaging

Artificial intelligence (AI) technology includes deep learning and machine learning algorithms.1,2 Like all fields, neurosurgery has become an increasing focus of research. However, AI is not yet clinically active in neurosurgery. This is the first case in the literature of a patient whose glioma was detected by an AI algorithm during ongoing magnetic resonance imaging (MRI), followed by an operation the next day at the Gazi University Faculty of Medicine Department of Neurosurgery.

The first use of AI technology in medicine was introduced into the literature by Gunn in 1976.3 The interest in medical AI has increased in the past two decades. The search for more accurate diagnoses, safe predictions of outcomes, and rapid treatments has triggered the development of medical AI.

Neurosurgeons have always been enthusiastic about using advanced technologies, including AI.4 For this reason, it is necessary to participate in developing AI technology. AI studies are promising, especially in the various subbranches of neurosurgery such as trauma, functional neurosurgery, and spine surgery. Although each branch has different goals and methods, all of the technological advancements share common goals, such as improving patient outcomes, helping physician decision making, and saving valuable time for both patients and healthcare workers.

Since January 2020, funded by the Presidency of the Republic of Türkiye, our institution has been the major participant of the Turkish Brain Project, which focuses on the development of AI-based brain imaging technologies. Within this project, a data set was developed to train a region-based convolutional neural network object-detection algorithm, and it is integrated into our hospital network, which processes MRI images in real time.

Illustrative Case

In October 2022, this algorithm alerted our research team with a detection. A 46-year-old female patient presented with a nonspecific headache and was admitted to undergo MRI. Physicians who received the alert examined the images and detected an object that was consistent with a diffuse glioma (Fig. 1). This happened while the patient was in the MRI scanner for additional sequences. Typically, it would take days for radiology to examine and report the results. An even longer time can pass before a patient’s results are evaluated. Meanwhile, the tumor can progress and cause permanent neurological deficits. Our patient was informed about the situation and admitted for a biopsy. A day later, the first patient diagnosed by an AI algorithm running on real-time data was operated on. The patient underwent a stereotactic biopsy operation. The pathology report confirmed an isocitrate dehydrogenase wild-type diffuse glioma. The patient was referred to the oncology department for evaluation and immediate treatment.

FIG. 1.
FIG. 1.

Axial fluid-attenuated inversion recovery images of the patient (A) and segmented images generated by AI (B).

Discussion

Observations

AI applications for neurosurgical practice have gained speed in the last 5 to 10 years.5 Lesion characterization via endoscopic or direct visualization in particular is among the most recently studied topics.6,7 This class of applications has the ability to detect subtle changes that could elude the evaluating physician and can note patterns that may or may not be visible to the eye.8 Radiomic studies that can predict molecular diagnoses or possible behaviors of a lesion are especially benefiting from this technology.9 Evaluation of histopathological specimens with the aid of AI is another hot topic in this area. It has been shown that glioma classifications can be identified with a precision greater than 85%10 or that the survival of a patient with glioblastoma can be predicted11 from a single hematoxylin and eosin–stained specimen. Furthermore, AI has demonstrated a use even in psychological evaluations and psychotherapy.12

Radiological image evaluation for a reliable diagnosis is still in its early stages. Although many different studies have shown that a reliable neurosurgical diagnosis could be made for different scenarios, such as subdural hematoma,13 traumatic brain injury,14 tumor detection,15 and classification,16 most of these studies are experimental. There are initiatives to create open-source17 or commercial software for automated MRI interpretation. In a literature search, there were reports on one of the commercial systems which indicates that the use of a similar computed tomography (CT)-based system reduces the scan view delay18 as well as the hospital length of stay for pulmonary embolism and intracranial hemorrhage.19 However, these results were obtained by comparing retrospective and prospective data, which carries the risk of unintentional bias associated with it. We were unable to identify any reports about a dedicated neurosurgical MRI processing system that is currently in use for everyday neurosurgical clinical practice in real time, such as ours. Most of these previously mentioned products are still under evaluation for neurosurgical pathologies, except for the CT-based intracranial hemorrhage triage system, and are not in active use for now. Prompt implementation of such technologies into standard clinical practice may be beneficial for patients and physicians.

Early resection is currently considered the best case management option, and early chemoradiotherapy is indicated in some cases according to the risk profile.20 With this in mind, shortening the period from MRI acquisition to surgery is beneficial for the patients. Despite the lack of available data, relatively recent studies have suggested that the median time from radiological diagnosis to first-line treatment is approximately 3 months.21 With the aid of this technology, we can reduce this time to days, as seen in this report. It is necessary to conduct further investigations to study the advantages of earlier treatment.

Lessons

We believe the time has come for the integration of AI in clinical practice. Alert systems such as ours will allow efficient and quick screening of complex real-life medical data. This allows an accurate and prompt prioritization of patients based on malignancy and the urgency of their diagnosis. Consequently, AI helps to save valuable time in patient care and improve healthcare quality and outcomes.

Acknowledgments

The authors thank the Digital Transformation Office of the Presidency of Türkiye for its substantial contribution to the development of the artificial intelligence model used through the Turkish Brain Project.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Celtikci, Borcek, Karaaslan. Acquisition of data: Celtikci, Sozer, Sagiroglu, Demirtas, Karaaslan. Analysis and interpretation of data: Celtikci, Sagiroglu, Karaaslan. Drafting the article: Celtikci, Sozer, Borcek, Sagiroglu, Poshtkouh, Demirtas. Critically revising the article: Celtikci, Borcek, Sagiroglu, Poshtkouh, Kuzucu. Reviewed submitted version of manuscript: Celtikci, Sozer, Sagiroglu, Poshtkouh, Kuzucu. Approved the final version of the manuscript on behalf of all authors: Celtikci. Statistical analysis: Celtikci. Administrative/technical/material support: Celtikci, Borcek, Karaaslan. Study supervision: Celtikci, Sagiroglu, Kuzucu.

References

  • 1

    Senders JT, Staples PC, Karhade AV, et al. Machine learning and neurosurgical outcome prediction: a systematic review. World Neurosurg. 2018;109:476486.e1.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Tewarie IA, Hulsbergen AFC, Gormley WB, Peul WC, Broekman MLD. Artificial intelligence in clinical neurosurgery: more than machinery. World Neurosurg. 2021;149:302303.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Gunn AA. The diagnosis of acute abdominal pain with computer analysis. J R Coll Surg Edinb. 1976;21(3):170172.

  • 4

    Panesar SS, Kliot M, Parrish R, Fernandez-Miranda J, Cagle Y, Britz GW. Promises and perils of artificial intelligence in neurosurgery. Neurosurgery. 2020;87(1):3344.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5

    Schilling AT, Shah PP, Feghali J, Jimenez AE, Azad TD. A brief history of machine learning in neurosurgery. Acta Neurochir Suppl (Wien). 2022;134:245250.

  • 6

    Harsono AB, Susiarno H, Suardi D, et al. Cervical pre-cancerous lesion detection: development of smartphone-based VIA application using artificial intelligence. BMC Res Notes. 2022;15(1):356.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Yang XX, Li Z, Shao XJ, et al. Real-time artificial intelligence for endoscopic diagnosis of early esophageal squamous cell cancer (with video). Dig Endosc. 2021;33(7):10751084.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Abe S, Oda I. Real-time pharyngeal cancer detection utilizing artificial intelligence: journey from the proof of concept to the clinical use. Dig Endosc. 2021;33(4):552553.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Bijari S, Jahanbakhshi A, Hajishafiezahramini P, Abdolmaleki P. Differentiating glioblastoma multiforme from brain metastases using multidimensional radiomics features derived from MRI and multiple machine learning models. BioMed Res Int. 2022;2022:2016006.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Jose L, Liu S, Russo C, et al. Artificial intelligence-assisted classification of gliomas using whole-slide images. Arch Pathol Lab Med. Published online November 29, 2022. doi:10.5858/arpa.2021-0518-OA.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Zadeh Shirazi A, Fornaciari E, Bagherian NS, Ebert LM, Koszyca B, Gomez GA. DeepSurvNet: deep survival convolutional network for brain cancer survival rate classification based on histopathological images. Med Biol Eng Comput. 2020;58(5):10311045.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Gual-Montolio P, Jaén I, Martínez-Borba V, Castilla D, Suso-Ribera C. Using artificial intelligence to enhance ongoing psychological interventions for emotional problems in real- or close to real-time: a systematic review. Int J Environ Res Public Health. 2022;19(13):7737.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Colasurdo M, Leibushor N, Robledo A, et al. Automated detection and analysis of subdural hematomas using a machine learning algorithm. J Neurosurg. 2022;138(4):10771084.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Hibi A, Jaberipour M, Cusimano MD, et al. Automated identification and quantification of traumatic brain injury from CT scans: are we there yet? Medicine (Baltimore). 2022;101(47):e31848.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Saravanan S, Kumar VV, Sarveshwaran V, Indirajithu A, Elangovan D, Allayear SM. Computational and mathematical methods in medicine glioma brain tumor detection and classification using convolutional neural network. Comput Math Methods Med. 2022;2022:4380901.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Chitnis S, Hosseini R, Xie P. Brain tumor classification based on neural architecture search. Sci Rep. 2022;12(1):19206.

  • 17

    Lachinov D, Vasiliev E, Turlapov V. Glioma segmentation with cascaded Unet. arXiv. Preprint posted online October 9, 2018. doi:10.48550/arXiv.1810.04008

  • 18

    Ginat D. Implementation of machine learning software on the radiology worklist decreases scan view delay for the detection of intracranial hemorrhage on CT. Brain Sci. 2021;11(7):832.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19

    Petry M, Lansky C, Chodakiewitz Y, Maya M, Pressman B. Decreased hospital length of stay for ICH and PE after adoption of an artificial intelligence-augmented radiological worklist triage system. Radiol Res Pract. 2022;2022:2141839.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Munkvold BKR, Solheim O, Bartek J Jr, et al. Variations in the management of diffuse low-grade gliomas—a Scandinavian multicenter study. Neurooncol Pract. 2021;8(6):706717.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21

    Obara T, Blonski M, Brzenczek C, et al. Adult diffuse low-grade gliomas: 35-year experience at the Nancy France neurooncology unit. Front Oncol. 2020;10:574679.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • Collapse
  • Expand
  • FIG. 1.

    Axial fluid-attenuated inversion recovery images of the patient (A) and segmented images generated by AI (B).

  • 1

    Senders JT, Staples PC, Karhade AV, et al. Machine learning and neurosurgical outcome prediction: a systematic review. World Neurosurg. 2018;109:476486.e1.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Tewarie IA, Hulsbergen AFC, Gormley WB, Peul WC, Broekman MLD. Artificial intelligence in clinical neurosurgery: more than machinery. World Neurosurg. 2021;149:302303.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Gunn AA. The diagnosis of acute abdominal pain with computer analysis. J R Coll Surg Edinb. 1976;21(3):170172.

  • 4

    Panesar SS, Kliot M, Parrish R, Fernandez-Miranda J, Cagle Y, Britz GW. Promises and perils of artificial intelligence in neurosurgery. Neurosurgery. 2020;87(1):3344.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5

    Schilling AT, Shah PP, Feghali J, Jimenez AE, Azad TD. A brief history of machine learning in neurosurgery. Acta Neurochir Suppl (Wien). 2022;134:245250.

  • 6

    Harsono AB, Susiarno H, Suardi D, et al. Cervical pre-cancerous lesion detection: development of smartphone-based VIA application using artificial intelligence. BMC Res Notes. 2022;15(1):356.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Yang XX, Li Z, Shao XJ, et al. Real-time artificial intelligence for endoscopic diagnosis of early esophageal squamous cell cancer (with video). Dig Endosc. 2021;33(7):10751084.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Abe S, Oda I. Real-time pharyngeal cancer detection utilizing artificial intelligence: journey from the proof of concept to the clinical use. Dig Endosc. 2021;33(4):552553.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Bijari S, Jahanbakhshi A, Hajishafiezahramini P, Abdolmaleki P. Differentiating glioblastoma multiforme from brain metastases using multidimensional radiomics features derived from MRI and multiple machine learning models. BioMed Res Int. 2022;2022:2016006.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Jose L, Liu S, Russo C, et al. Artificial intelligence-assisted classification of gliomas using whole-slide images. Arch Pathol Lab Med. Published online November 29, 2022. doi:10.5858/arpa.2021-0518-OA.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Zadeh Shirazi A, Fornaciari E, Bagherian NS, Ebert LM, Koszyca B, Gomez GA. DeepSurvNet: deep survival convolutional network for brain cancer survival rate classification based on histopathological images. Med Biol Eng Comput. 2020;58(5):10311045.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Gual-Montolio P, Jaén I, Martínez-Borba V, Castilla D, Suso-Ribera C. Using artificial intelligence to enhance ongoing psychological interventions for emotional problems in real- or close to real-time: a systematic review. Int J Environ Res Public Health. 2022;19(13):7737.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Colasurdo M, Leibushor N, Robledo A, et al. Automated detection and analysis of subdural hematomas using a machine learning algorithm. J Neurosurg. 2022;138(4):10771084.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Hibi A, Jaberipour M, Cusimano MD, et al. Automated identification and quantification of traumatic brain injury from CT scans: are we there yet? Medicine (Baltimore). 2022;101(47):e31848.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Saravanan S, Kumar VV, Sarveshwaran V, Indirajithu A, Elangovan D, Allayear SM. Computational and mathematical methods in medicine glioma brain tumor detection and classification using convolutional neural network. Comput Math Methods Med. 2022;2022:4380901.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Chitnis S, Hosseini R, Xie P. Brain tumor classification based on neural architecture search. Sci Rep. 2022;12(1):19206.

  • 17

    Lachinov D, Vasiliev E, Turlapov V. Glioma segmentation with cascaded Unet. arXiv. Preprint posted online October 9, 2018. doi:10.48550/arXiv.1810.04008

  • 18

    Ginat D. Implementation of machine learning software on the radiology worklist decreases scan view delay for the detection of intracranial hemorrhage on CT. Brain Sci. 2021;11(7):832.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19

    Petry M, Lansky C, Chodakiewitz Y, Maya M, Pressman B. Decreased hospital length of stay for ICH and PE after adoption of an artificial intelligence-augmented radiological worklist triage system. Radiol Res Pract. 2022;2022:2141839.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Munkvold BKR, Solheim O, Bartek J Jr, et al. Variations in the management of diffuse low-grade gliomas—a Scandinavian multicenter study. Neurooncol Pract. 2021;8(6):706717.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21

    Obara T, Blonski M, Brzenczek C, et al. Adult diffuse low-grade gliomas: 35-year experience at the Nancy France neurooncology unit. Front Oncol. 2020;10:574679.

    • PubMed
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 1227 1227 49
PDF Downloads 1073 1073 52
EPUB Downloads 0 0 0