Purpose: Oral squamous cell carcinoma (OSCC) is the most common oral malignancy, representing approximately 90% of all oral cavity cancers. In consideration of the complexity of the oral cavity, imaging is integral to the classification of the primary tumor and disease extent. We propose the use of three-dimensional ultrasound (3DUS) as a point-of-care imaging modality for bedside oral cancer identification and tumor volume quantification. This work presents the clinical utility of 3DUS in a clinical case study of OSCC compared to healthy control.
Methods: We developed a motorized handheld 3DUS device housing custom-printed 14L5 (10MHz) and 11L3 (7MHz) linear transducer-holders. Trans-submandibular 3DUS acquisition protocol involved 5.0cm linear translation at 4.0cm depth-of-field by a trained observer. 3DUS images of the patient and healthy control were acquired and anatomical structures were identified. In the pathological case, the maximum tumor dimensions in anteroposterior (AP) and craniocaudal (CC) planes and depth-of-invasion (DOI) were calculated and compared with preoperative CT and MRI measurements by quantifying the error in tumor volume estimation.
Results: Bilateral visualization was attained of the oral floor anatomy with 10 MHz and 7 MHz transducers in both healthy and pathological cases. Preoperative CT and MRI images indicated tumor dimensions: 1.60x1.12mm (AP/CC) and 0.44cm DOI, and 1.45x1.90 cm (AP/CC) and 0.66cm DOI. 3DUS with the 10MHz transducer allowed for the approximation of tumor measurements at 1.44x1.37cm (AP/CC) with a 0.49cm DOI. The error in dimension estimation was 6%, 9%, and 11% for AP, CC, and DOI between the 3DUS and mean preoperative dimensions between CT and MRI images.
Conclusion: We demonstrated clinical utility of 3DUS in OSCC in the oral floor. 3DUS acquisition enabled the complete visualization of the oral cavity anatomy, which supplants the demand for preoperative bedside imaging of the primary site and tumor dimension estimation.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), the Canadian Institutes of Health Research (CIHR), and the Ontario Institute for Cancer Research (OICR) Imaging Program.