Radiologists are putting patients at the center of care with the use of 3D modeling in surgical treatment planning.  By leveraging image overlay tools available on FUJIFILM Medical’s Synapse 3D solution, radiologists at Keck Hospital, USC Norris Comprehensive Cancer Center and Hospital, Los Angeles, are using volumetric imaging to generate 3D models of organs and other parts of the anatomy. Referring physicians and radiologists are using these 3D models to develop treatment plans, predict and prepare for complications during procedures and educate patients about their conditions prior to treatment.

Today, 3D models already have several clinical applications across the continuum of care, including creating prosthetics, individualized bio-printing of organs, transplantation and transcatheter aortic valve replacement planning.¹  3D prototypes can be rendered digitally for viewing on a monitor, overlaid on video or printed out as physical models. With video overlay, a segment of a video recording can be registered to the corresponding preoperative 3D-computed tomography (CT) image.²

To develop a printed model, volumetric image datasets—such as CT or MRI—are segmented into structures of interest. These structures are then tessellated and outputted in the Standard Tessellation Language (STL) file format, a universally accepted format for 3D rendering and printing. An ever-growing variety of 3D printers can take the STL files and print a three dimensional physical object.

The use of 3D models in medical imaging is a logical extension of traditional image post-processing techniques. Fujifilm Medical supports clinicians in developing image datasets into models with Synapse 3D, an advanced visualization solution for multimodality image processing.  Fujifilm Medical’s Synapse 3D Clinical Application Suite is integrated into Synapse PACS, where the volumetric image datasets reside.

Synapse 3D includes a comprehensive base toolset, complete with automatic vessel segmentation and analysis algorithms, one-click measurement capabilities, masking segmentations, sector multiplanar reformations (MPR) and a fusion overlay tool and fusion viewer, enabling radiologists to overlay and co-register a series of images. Real-time collaboration between radiologists and referring physicians is supported through screen sharing technology on a 3D/4D viewer, where radiologists and referrers can manipulate and evaluate images across multiple planes using Synapse 3D.

3D models reshape renal cancer treatment

This year alone, renal cell and renal pelvis cancer have accounted for an estimated 63,920 new cases and 13,860 deaths in the United States.3 One of the standards of care for treating small localized renal tumors is robot-assisted laparoscopic partial nephrectomy, and preoperative imaging techniques, such as CT or MRI, are routinely performed to determine the tumor location, dimensions and charecterization.2 Increasingly, radiologists are leveraging image overlay techniques to create 3D models to help surgeons prepare for laparoscopic partial nephrectomy. Using these models, surgeons can track the kidney surface in real-time when applied to intraoperative video recordings and overlay the 3D models of the kidney, tumor (or stone) and collecting system.4

At Keck Hospital, USC Norris Comprehensive Cancer Center and Hospital, Los Angeles, radiologists scan an estimated 400 patients annually who present for a consultation for renal cancer surgery; around 300 of those patients eventually undergo surgery.  Radiologists and surgeons use Synapse 3D to develop virtual 3D models for the evaluation of renal masses and to develop treatment plans.

Radiologists use standard protocols in Synapse 3D to enhance the visualization of the mass and extract information that the urologist then uses. 

Synapse 3D allows the radiologist to overlay and fuse the images from phases one, two, three and four. Users can subtract the blood vessels or the collecting system from the kidney to better view the functioning parenchyma or blood vessels.  Different visualization streams can be configured to see everything, extract features or superimpose another, making the tumor transparent.

Case study no. 1

A recent patient presented with pain in the abdomen and blood in his urine.  After undergoing a single-phase CT scan at another center, he was referred to physicians at the USC Norris Comprehensive Cancer Center and Hospital, where he underwent a standard of care multiphase-CT scan. In this particular case, we looked at the prior CT scan, but evaluated the multiphase-CT scan of the tumor on Synapse 3D, using all of the phases in Synapse. The virtual 3D model was created by registering all four volume datasets.

The imaging findings showed a 6-cm mass based in the right kidney, which extended into the renal sinus. This confirmed that the mass was causing hematuria, as well as pain for the patient. 

Figure 1. (Left) A 74-year-old male with a mass at the upper pole of the right kidney. The mass is seen in the nephrographic phase of the CT (arrow) extending into the renal sinus and hilum. (Right) A virtual model using all the phases in the CT scan helps demonstrate the relationship between the tumor (purple), renal cortex (translucent light purple), the collecting system of the kidney (white) and the arteries supplying the kidney.

After measuring the remainder of the kidney and the size of the tumor, we estimated— working with the urologists—what the surgical plane would be and developed a treatment plan. Prior to Synapse 3D, the same process was very labor intensive; now that it has been optimized, it can be done on the fly at our workstations.

A critical step in developing a successful surgical plan takes place when the radiologist and urologist review the images together in real time. The Synapse workstation supports screen sharing and universal access to the advanced visualization tools; because the 3D tools are native in the PACS, the physicians can manipulate the images and evaluate them from multiple angles.

The radiologist reviews the scan with the urologist, who determines the surgical plane, giving the surgeon a better idea of what structures need to be resected, what possible complications might arise and what to prepare for in the surgery. Models of the vascularity and specific road maps are created and sent to PACS and the electronic medical record for the surgeon’s review.

Surgeon prepares for the procedure

In preparing for laparoscopic partial nephrectomy, the surgeon must consider several questions.  These may include whether there are additional blood vessels, if the tumor has spread to the surrounding tissue, whether there are additional anatomic considerations or if the kidney is very low lying, malrotated or in a different position. The 3D models allow the surgeons to anticipate if they may have to resect the collecting system and prepare additional materials to stop urinary leaks.

The renal team has developed algorithms using the 3D models to help predict complications that may arise during the procedure, how long it might take and whether or not the patient will need a blood transfusion during surgery.  If surgeons know there is going to be an additional artery supplying the kidney, they will know to look for it. The more information the surgeons have, the better prepared they are for possible complications. 

In certain masses, the gold standard is to resect a portion of the kidney, but sometimes the tumor is very large or has infiltrated the entire kidney, in which case the surgeon will need to resect the entire kidney. The 3D models help the surgeons decide what surgery they are going to do before they make an incision in the patient. In this case, the patient had a partial robotic nephrectomy.

The virtual 3D model also facilitates patient education. The surgeon in this case used the model to explain to the patient where the tumor was, what was causing the bleeding and pain and identifying the area where the mass would be resected and the part of the kidney that would be lost.

During this surgery, the surgeons used the 3D digital models viewed on large monitors as a guide.  Many surgeons use the 3D models not just to plan the surgery, but as a reference during surgery.

Currently, the radiology department and USC Urology Institute Chair Inderbir S. Gill, MD, are collaborating with Jean Christophe Bernhard, MD, a visiting urologist from Bordeaux, France, to create these physical models.

Diagnostic confidence: Case study no. 2

Another important benefit of generating 3D models is the confidence that it instills in the radiologist when making a diagnosis. In a recent case, a 78-year-old male patient presented with diarrhea and vomiting, and images from a multiphase-CT study showed an incidental finding of a renal mass (Figure 2).

Such an incidental finding is not uncommon: approximately 75% of all renal masses that go to surgery are identified incidentally.  In this case, the patient had multiple comorbidities, so the questions raised were how much of a risk the mass posed to the patient’s health and how much the patient would benefit from surgery.

Figure 2. (Left and Middle) A 78-year-old man with an incidental finding of a mass in the left upper pole, demonstrated as a cortical-based lesion. (Right) A virtual model of the mass shows its relationships to the rest of the kidney. Evaluation using Synapse 3D determined that the mass grew by 4 mm in a period of 42 months. Considering the patient’s comborbidities, a decision was made to continue monitoring the lesion with imaging.

Looking at the nature of the mass and imaging appearances, we determined that it was a relatively low-grade tumor. Synapse 3D gave us the diagnostic confidence that we needed to determine that it was a low-grade tumor for the following reasons: one, it didn’t enhance very much; two, it was fairly uniform; and three, its shape was small and round. The virtual 3D models helped in conveying this information to the surgeon. 

Since the renal mass was not causing any symptoms and the patient had several comorbidities, he was unlikely to benefit from surgery.  The surgeon agreed with the radiologist’s recommendation to monitor the tumor with an imaging study every six months.   The surgeon then explained the treatment plan to the patient using the virtual 3D model.

In this case, the radiologists at USC  will follow up with imaging every six months and will use Synapse 3D to compare the prior and current scans to look for changes in the tumor.

Conclusion

While radiologists have worked with 3D volumetric image data for many years, they did not have the technology to present the data in a practical and relatable manner. That has changed with solutions like Synapse 3D. 

These tools make it practical for radiologists to increasingly move away from 2D toward 3D imaging, especially since much of the image data is already acquired as a 3D volume. If we don’t use 3D, we are wasting information. In the past, we lacked the tools, but today, with Synapse 3D and the virtual 3D models that we create, radiologists can extract and extrapolate more information into forms that are useful for physicians for multiple purposes.

Vinay A. Duddalwar, MD, FRCR, is abdominal imaging section chief and medical director of imaging, Keck Hospital, USC Norris Comprehensive Cancer Center and Hospital in Los Angeles.

References

1. Rengier F, Mehndiratta A, von Tenngg-Kobligk H, et al. 3D printing based on imaging data: review of medical applications. Int J Comput Assist Radiol Surg. 2010 Jul;5(4):335-41. 
2. Vagvolgyi BP, Reiley CE, Hager GD, Taylor RH.  Augmented reality using registration of 3D computed tomography to stereoscopic video of laparscopic renal surgery. Johns Hopkins Medical Institutions, Brady Urological Institute, Baltimore MD. AUA Annual Meeting. May 17 – 22, 2008. Abstract 689.
3. Kidney Cancer. National Cancer Institute. http://www.cancer.gov/cancertopics/types/kidney. Accessed September 25, 2014.
4. S Li-Ming, BP Vagvolgyi, Agarwal R, Reiley CE, et al. Augmented Reality During Robot-assisted Laparoscopic Partial Nephrectomy: Toward Real-Time 3D-CT to Stereoscopic Video Registration. J Urology. 2008. DOI: 10.1016/j.urology.2008.11.040