Clinical.
We present the results of the first 16 volunteers who have been recruited and imaged under a protocol approved by the Ethical Committee of Regione Umbria,
Italy (N.
6845/15/AV/DM of 14/10/2015,
N.
10352/17/NCAV of 16/03/2017).
The protocol concerns a feasibility study for detection of breast cancers using the proposed microwave mammogram apparatus,
with the aim of quantifying the potential of the proposed microwave mammogram apparatus to be used for screening.
The informed consent was obtained from all volunteers; moreover,
4 volunteers underwent the microwave imaging for the two breasts,
and 12 volunteers underwent the microwave imaging for just one breast.
All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
The radiologist study reviewed the conventional exam for each volunteer; conventional exam was constituted by echography and/or mammography or (limited to one case) magnetic resonance imaging.
Echography was performed using the MyLab 70 xvg Ultrasound Scanner (Esaote,
Genova,
Italy); mammography was performed using Selenia LORAD Mammography System (Hologic,
Marlborough,
MA); magnetic resonance imaging was performed through a 3T scanner (Siemens Healthcare,
Erlangen,
Germany).
The output of the radiologist study review is listed in Table 1,
together with subjects’ details.
Specifically,
where possible,
the breast type has been classified according to its density,
following the scale defined by the American College of Radiology (ACR) which goes from ACR1 (extremely fatty breast) to ACR4 (extremely heterogeneous fibroglandular breast) [9].
The inclusion type,
if present,
has been classified according to [10-12].
Once a subject agrees to participate in the protocol,
the clinical study coordinator assists the subject with the positioning of her breast in the microwave apparatus.
The exam data acquisition is completely computer controlled under the observation of a microwave system operator present in the room.
Once the exam is completed,
the data is transferred to a secure server for imaging.
Microwave Apparatus.
All microwave in-vivo images and results shown in this paper were reconstructed from data gathered using the apparatus installed at the Department of Diagnostic Imaging,
Perugia Hospital,
Perugia,
Italy.
The apparatus is constituted by one transmitting antenna (denoted here as TX) and by one receiving antenna (here denoted as RX).
Both antennas operate in air,
in the frequency band of 1 to 9 GHz.
The apparatus is additionally constituted by: a hub with a cup that are placed to contain the breast of the patient (prone positioned) and two arms to rotatably associate TX and RX to the hub.
The TX antenna is placed more radially external compared to the RX antenna.
The RX antenna is placed more radially external with respect to the cup containing the breast.
Both TX and RX antenna are configured to be rotated around the azimuth,
such that they can pick up the reflected electromagnetic field in all the different directions.
Both antennas are connected to a Vector Network Analyzer VNA (Copper Mountain Technologies,
IN,
USA),
which uses an output power of 1mW.
Some details of the apparatus (appropriately integrated in a bed) are shown in Fig 2.
For each TX and RX position,
S21 was acquired at 1601 frequencies from 1 to 9 GHz in 5 MHz increments.
TX and RX were positioned at the same height on the azimuth plane which crosses the centre of the breast of the prone subject (after checking that the half power beam angle of the antennas include the breast).
On such azimuth plane,
we used 15 transmitting positions,
divided in 5 groups centred at 0°,
72°,
144°,
216°,
288°; each group contains 3 transmitting positions displaced 4.5° from each other.
Moreover,
we used 80 receiving positions displaced 4.5° from each other.
Microwave images.
To generate the image,
the signals measured by the receiving antenna,
i.e.
the S21 output from the VNA,
are processed by a processing unit through an imaging algorithm based on HP.
We reconstructed the two-dimensional images in the azimuthal plane,
i.e.
coronal plane.
In more details,
we reconstructed images in a cylindrical grid with radius of 7 cm (equal to the receiving antenna radius).
All images were acquired by employing (off-line) the imaging algorithm.
To be able to perform intra-breasts comparison,
all intensity images have been normalized to unitary average.
For each microwave image,
we calculated the parameter Max/Avg (maximum of intensity divided by the average of intensity).
Based on the classification performed by the radiologist,
we calculated: the mean and standard deviation of Max/Avg for the healthy breasts; the mean and standard deviation of Max/Avg for the non-healthy breasts.
We denoted with T the sum of mean and standard deviation of Max/Avg for the healthy breasts.
Next,
T has been used to classify the microwave images of non-healthy breasts.
Specifically: if Max/Avg ≥ T the image was classified as altered,
while if Max/Avg < T the image was classified as non-altered.